Search

Anomaly detection in photovoltaic (PV) cells is crucial for ensuring the efficient operation of solar power systems and preventing potential energy losses. In this paper, we propose an enhanced YOLOv7-based deep learning framework for fast and accurate anomaly detection in PV cells. Our approach incorporates Partial Convolution, Switchable Atrous Convolution and novel data augmentation techniques to address the challenges of varying defect sizes, complex backgrounds. The Partial Convolution component effectively manages the irregularities in PV cell images, reducing false detections. On the other hand, the Switchable Atrous Convolution enhances the model's adaptability to different defect scales and spatial resolutions, leading to improved localization and classification performance. We evaluate our model on a large-scale dataset of PV cell images, demonstrating its superiority over existing methods in terms of detection accuracy, speed, and robustness. Our proposed framework offers a practical and reliable solution for real-time anomaly detection in PV cells, facilitating timely maintenance and maximizing the performance of solar energy systems. The integration of these advanced convolution techniques into the YOLOv7 model not only improves detection capabilities but also paves the way for further research and development in the field of deep learning-based anomaly detection. • The PSA-YOLOv7 proposed in this paper exhibits better interpretability than the baseline model. • The PSA-YOLOv7 shows better robustness under common corruptions. • PSA-YOLOv7 has improved anomaly detection performance while ensuring real-time capability. [ABSTRACT FROM AUTHOR]

The objective of this paper is to optimize the energy consumption and performance of fuel cell electric bicycles (FCEBs) under specific key input parameters. The paper applied an integrated method that includes an artificial neural network (ANN) and genetic algorithm (GA) to forecast and identify an optimal performance and energy consumption of FCEBs. The simulation model of FCEBs is established and simulated in MATLAB-Simulink environment to generate 1000 data points, that are used for training, validating, testing artificial neural network, the ANN architecture containing five input neurons, two hidden neurons and two output neurons, respectively. Furthermore, the GA is integrated to find the maximum performance and energy consumption once the ANN has exactly been trained. The study found that the FCEB configuration can achieve an effective performance at 30.3 km/h with required power of 210.4 W under speed level_5, radius of wheel 0.39 m, frontal area 0.423 m2, slope grade 0%. In order to validate and verify the simulated results, the experimental approach method was conducted in the same condition. The experimental results fit well with the simulated results in the same initial input parameters. • A method integrating the artificial neural network with genetic algorithm is applied. • The consumed power and fuel cell voltage are optimized by combining artificial neural network and genetic algorithm (ANN-GA). • The effective performance area of fuel cell electric bicycle under various a structure and operating parameters is identified. • Deep-learning results reflect major physical mechanism of fuel cell electric bicycle operation. [ABSTRACT FROM AUTHOR]

The motivation of this paper is to investigate the resistance of countries' energy research and development (RD&D) expenditures to random shocks. The analysis includes the five countries (France, Germany, Japan, the United States, and the United Kingdom) that are the biggest investors in RD&D in fossil fuels, renewables, energy efficiency, and nuclear energy. Thus, economies will be able to take precautions against global shocks in producing sustainable energy policies. To achieve this aim, this paper uses the artificial neural network approach, which uses a distributed computing model that can store and generalize data after a learning period. Empirical results widely depend on countries and distinct energy technology policy fields. The key findings provide evidence that RD&D expenditures, except for renewables, do not tend to mean revert. The economies of Japan, Germany, and the United States should make more renewable investments to reduce the resistance of renewable energy sources to shocks and thus produce policies for both environmental sustainability and energy security. The economies of France and the United Kingdom can ensure energy security by continuing their sustainable energy policies. • Applied SDG policies focused on energy technology RD&D are presented. • The stochastic characteristics of energy technology investments are examined. • Carbon-reducing policies are recommended for the top five economies. • Energy technologies are perused through artificial intelligence approach. [ABSTRACT FROM AUTHOR]

Integrating enormous quantities of wind energy into the electrical system requires precise planning and forecasting. This paper presents a novel framework for wind power forecasting, which establishes a new standard for accuracy and reliability. It uses advanced deep neural networks and a stacking ensemble mechanism to make probabilistic forecasts. These forecasts address wind speed volatility and variability, which continue to be key issues in producing consistently accurate wind power forecasts despite advances in deep learning. Base learners generate forecasts that include lower and upper bounds, as well as median prediction intervals. The Huber regressor, also known as the Meta-learner, combines projections from many models to reduce susceptibility to extreme values. Expanding window cross-validation is used in performance evaluation, with a focus on Mean Absolute Error, Mean Squared Error (MSE), Symmetric Mean Absolute Percentage Error, Prediction Interval Coverage Probability, and Average Interval Width for one, two, and three step ahead predictions. The model's stability is measured by the standard deviation of MSE throughout each validation window. Additionally, the Diebold–Mariano test is used to validate the Meta-learner's predictive accuracy in contrast to other advanced models and the seasonal naïve benchmark. This study found that the meta model outperformed all other models in eight of the nine forecasts tested, indicating its effectiveness in medium-range forecasting. In Germany, it exceeded Temporal Fusion Transformer by two and three steps ahead, with 22% and 34% improvements, respectively. Indeed, this study not only provides a reliable prediction tool, but also lays foundations for efficient and effective energy management and policy planning in the renewable energy industry. • Deep neural network architectures are applied to probabilistic wind power forecasting. • Transformer and artificial neural network approaches combined with conformal prediction intervals is proposed. • A window cross-validation scheme on wind power forecasting is evaluated. • Wind power data from Germany, the UK, and the Netherlands are used for forecasting purposes. • SHAP and LIME analyses perform the explanatory analysis of the variables in the meta-learner's predictions. [ABSTRACT FROM AUTHOR]

Indirect evaporative cooling has garnered considerable attention and research owing to its notable advantages, especially high cooling efficiency and low energy consumption. However, the complexity and diversity of parameters affecting indirect evaporative cooler (IEC) performance necessitate the development of a fast and efficient performance prediction model, which remains a critical research objective. To address the particular challenge, this paper introduces a novel approach that utilizes the principles of neural network theory to construct a predictive model based on backpropagation artificial neural networks (BP-ANN). The developed model is employed to investigate the impact of six main operating parameters of the water spray system on the IEC performance. Wet-bulb efficiency and pressure drop of secondary air channel are utilized as the performance evaluation indexes. The model is rigorously validated by relevant experimental data, enabling a comprehensive analysis of the influence of each parameter on the operational performance of the whole system. The significance of the influential parameters is evaluated through grey relational analysis. Furthermore, a multi-objective optimization approach employing a genetic algorithm (GA) is introduced to effectively determine the optimal operating parameters of the IEC system. The results demonstrate that the whole system can efficiently operate with low energy consumption when the wet-bulb efficiency and pressure drop weights are balanced. The optimal operating parameters include a nozzle pressure of 0.15 MPa, spacing of 85 mm, aperture of 6 mm, count of 11, spray density of 0.85 kg/(s ∙ m2) and air velocity of 2 m/s, at which time the wet-bulb efficiency was 0.85, the pressure drop was 210 Pa, and the coefficient of performance (COP) was 12.4, which demonstrated an efficiency increase of 23.2% in comparison with the base system. The findings of this research significantly contribute to a comprehensive understanding of the influence of water-spray operating conditions on IEC performance, and provide valuable insights for the design and operation of high-efficient IEC systems in practical applications. • The datasets were generated by a laboratory-scale prototype of the IEC system. • The ANN model for IEC considering spray operation was trained and validated. • Effects of nozzle and operation characteristics parameters were investigated. • The significance of the influential parameters was evaluated. • The optimal parameters for IEC were determined based on genetic algorithms. [ABSTRACT FROM AUTHOR]

[Display omitted] • The effects of OEC on PN emissions of a gasoline engine are studied. • The first ANN model to predict the OEC gasoline engine performance is designed. • The NO x -ANN model is developed at 108 sets of experimental data. • The HC-CO-PN-T tq -ANN model is developed at 125 sets of experimental data. • Invalid NO x data are effectively predicted in the NO x -ANN model. The experiment about gasoline fuel plus oxygen-enriched combustion (OEC) technology is conducted on a combined injection engine in this paper. OEC is an effective in-cylinder clean combustion technology, but it will lead to extremely high NO x emissions and even cause the NO x detector failure at high oxygen ratios. The particulate number (PN) emissions on OEC are emphatically analyzed in this paper, which are not covered in previous studies. The results show that with the increase of oxygen ratio, PN emissions first increase and then decrease. After analyzing the experimental results, convenient and efficient artificial neural network (ANN) models are developed to predict the torque output, HC, CO, NO x , and PN emissions based on Levenberg-Marquardt, Bayesian regularization and Scaled conjugate gradient algorithms. The models have good fitting and prediction performance with correlation coefficient values > 0.99 and mean square error values less than 1E-3, which effectively predict the invalid NO x emissions data. The predicted results show that the ANN model can assist the bench test study as an effective tool for predicting the emissions and power performance of gasoline engines with OEC technology. [ABSTRACT FROM AUTHOR]

Rapid changes in the wind field of tropical cyclones can cause excessive loads and threaten the safety of offshore wind turbines. This paper designs a wind turbine yaw optimization strategy based on the deep neural network to reduce the structural loads of wind turbines caused by tropical cyclones. Firstly, the high-precision tropical cyclone data is used to analyze the characteristics of the wind field. Then, a pseudo-Monte Carlo experiment is designed to compensate for the incompleteness of the observed data. Furthermore, a robust nonlinear coupling model between the wind characteristics and the loads of the wind turbine is constructed by a deep neural network, and the optimal yaw angle is searched in real time based on this model. The simulation results show that the proposed deep neural network model based on pseudo-Monte Carlo scenario generation can robustly calculate the structural loads of wind turbines with a calculation error of less than 8 %. After applying this model to the real-time optimization control loop, the corresponding optimized yaw angle can be obtained according to the operating data of the wind turbine under tropical cyclone conditions so that the structural loads of the wind turbine are reduced. Compared with the no-yaw and traditional yaw strategies, the load suppression effect is 1 %–9 % under different working conditions. The proposed data-driven structural load model and load suppression algorithm will effectively improve the operating safety of wind turbines under tropical cyclone conditions. • Insufficient tropical cyclone observation data has been made up; • Deep neural network load models can be calculated quickly and accurately; • Optimizing yaw strategy reduces wind turbine structural loads in tropical cyclones. [ABSTRACT FROM AUTHOR]

Rapid changes in the wind field of tropical cyclones can cause excessive loads and threaten the safety of offshore wind turbines. This paper designs a wind turbine yaw optimization strategy based on the deep neural network to reduce the structural loads of wind turbines caused by tropical cyclones. Firstly, the high-precision tropical cyclone data is used to analyze the characteristics of the wind field. Then, a pseudo-Monte Carlo experiment is designed to compensate for the incompleteness of the observed data. Furthermore, a robust nonlinear coupling model between the wind characteristics and the loads of the wind turbine is constructed by a deep neural network, and the optimal yaw angle is searched in real time based on this model. The simulation results show that the proposed deep neural network model based on pseudo-Monte Carlo scenario generation can robustly calculate the structural loads of wind turbines with a calculation error of less than 8 %. After applying this model to the real-time optimization control loop, the corresponding optimized yaw angle can be obtained according to the operating data of the wind turbine under tropical cyclone conditions so that the structural loads of the wind turbine are reduced. Compared with the no-yaw and traditional yaw strategies, the load suppression effect is 1 %–9 % under different working conditions. The proposed data-driven structural load model and load suppression algorithm will effectively improve the operating safety of wind turbines under tropical cyclone conditions. • Insufficient tropical cyclone observation data has been made up; • Deep neural network load models can be calculated quickly and accurately; • Optimizing yaw strategy reduces wind turbine structural loads in tropical cyclones. [ABSTRACT FROM AUTHOR]

Building Performance Optimization (BPO) plays a pivotal role in enhancing building performance, guaranteeing comfort while reducing resource consumption. Existing performance-driven generative design is computational demanding and difficult to be generalized to other similar buildings with difficult to be generalized to other building types or climate conditions. To fill this gap, this paper introduces a novel framework, which integrates multitask learning (MTL), code compliance check, and multi-objective optimization through NSGA-III algorithm. This framework is able to identify Paratoo Optimal design solutions, which comply with building codes, at low computation costs. The framework begins with selecting key design variables that are critical to building energy, comfort performance and life cycle cost. It then employs MTL to enhance the model's accuracy while reducing computational costs. Next, we designed a code compliance checking module followed by the NSGA-III optimization process, with the objective of identifying solutions that comply with existing building codes. The results indicate that the proposed MTL network achieved an R2 score of 0.983–0.993 on the test set. In the particular case study where equal weights are preferred, this approach yielded noteworthy reductions of 27.65%, 19.55%, and 31.13% in Building Energy Consumption (BEC), Life Cycle Cost (LCC), and Residue of continuous Daylight Autonomy (RcDA), respectively, for a rural dwelling, and exclude solutions that fail to satisfy regulatory standards. This framework allows designer to input the weights of each objective based on their preference and can be applied to other building types and climate regions. Last, we develop a solution selection tool based on the results output by the framework we proposed, which can be found at https://github.com/LiMingchen159/Village-House-Design-Strategy-in-Hebei-Province-China. [Display omitted] • Introduced a novel framework for generative design. • Identified code-complied Pareto optimal designs in a computational efficient way. • Used Muti-task learning to enhance surrogate model accuracy. • Developed a tool for designers to tailor solutions by adjusting objective weightings. • Validated with a case study on passive design for rural homes in cold regions. [ABSTRACT FROM AUTHOR]

Improving the efficiency of home energy management (HEM) is of great significance in reducing resource waste and prompting renewable energy consumption. With the development of advanced HEM systems and internet-of-things technology, more residents are willing to participate in the electricity market through a demand response mechanism, which can not only reduce the electricity bill but also stabilize the operation of the main grid through peak shaving and valley filling. However, the uncertainties from household appliance parameters, user behaviors, penetrated renewable energy, and electricity prices render the HEM problem a non-trivial task. Although previous deep reinforcement learning (DRL) methods have no requirement for system dynamics due to their model-free and data-driven merits, the unrestricted action space may violate the physical constraints of various devices, and few works have concentrated on this area before. To solve the above challenges, this paper proposes a novel interior-point policy optimization-based multi-agent DRL algorithm to optimize the home energy scheduling procedure while guaranteeing safety during its operation. Besides, a time-series prediction model based on a non-stationary Transformer neural network is proposed to predict the future trend of solar generation and electricity price, which can provide the agents with abundant information to make better decisions. The superior performance of our proposed predictive-control coupled method is demonstrated through extensive numerical experiments, which achieves more precise prediction results, near-zero constraint violations, and better trade-offs between cost and safety than several benchmark methods. • A data-driven secure home energy management model under various uncertainties is proposed. • Multi-agent deep reinforcement learning framework based on deep deterministic policy gradient is established. • An interior-point optimization-based secure decision-making approach to conduct the home energy management process. • The uncertainties from photovoltaic and electricity price are solved by utilizing a novel non-stationary Transformer deep model. [ABSTRACT FROM AUTHOR]

Liquid cooling is a promising approach for battery thermal management due to its compact construction and high heat transfer coefficient. However, the conventional liquid cooling plates are generally designed by trial-and-error approaches regarding the battery as a uniform heat generating body during charge and discharge processes, limiting the cooling performance of the battery thermal management systems. This paper introduces a deep learning framework that considers non-uniform heat generation to design a liquid cooling plate with a bionic leaf structure. The framework links the structural features of the cooling plate to performance using deep learning and optimizes these features with a non-dominated sorting genetic algorithm II. The results show that the maximum temperature difference of the battery can be up to 11°C in natural cooling condition at 4C. The average value of of the three outputs of the established ANN neural network model is 0.98, and the model is capable of predicting results with high accuracy. More significantly, the temperature difference and pressure drop are 2.26°C and 2562.62 Pa compared to traditional structure, which are reduced by 60.17% and 25.06%, respectively. The proposed deep learning-based multi-objective optimization framework can guide the design of liquid cooling plates for battery thermal management systems. • A model considering the non-uniform heat generation characteristics is developed. • A novel bionic leaf structure liquid cooling plate is designed and optimized. • Deep learning is used to predict the performance of liquid cooling plates. • A multi-objective optimization framework combining deep learning and NSGA-II is developed. [ABSTRACT FROM AUTHOR]

Highlights • Exhaustive study of ANNs used in wind energy systems, identifying the most employed techniques. • ANNs classified into: forecasting and predictions; fault detection and diagnosis; optimal control; design optimization. • A statistical analysis of the current state (and future trends) in this field is done. Abstract Wind energy has become one of the most important forms of renewable energy. Wind energy conversion systems are more sophisticated and new approaches are required based on advance analytics. This paper presents an exhaustive review of artificial neural networks used in wind energy systems, identifying the methods most employed for different applications and demonstrating that Artificial Neural Networks can be an alternative to conventional methods in many cases. More than 85% of the 190 references employed in this paper have been published in the last 5 years. The methods are classified and analysed into four groups according to the application: forecasting and predictions; design optimization; fault detection and diagnosis; and optimal control. A statistical analysis of the current state and future trends in this field is carried out. An analysis of each application group about the strengths and weaknesses of each ANN structure is carried out. A quantitative analysis of the main references is carried out showing new statistical results of the current state and future trends of the topic. The paper describes the main challenges and technological gaps concerning the application of ANN to wind turbines, according to the literature review. An overall table is provided to summarize the most important references according to the application groups and case studies. [ABSTRACT FROM AUTHOR]

Highlights • A two-stage model is developed to quantify the uncertainty of PV power forecasts. • An integrated artificial neural network model is built to perform point forecast. • The prediction intervals (PI) associated to the point forecasts are evaluated. • The PIs cover more comprehensive information than the point forecasts. Abstract Due to the intermittency and uncertainty in photovoltaic (PV) power outputs, not only deterministic point predictions (DPPs), but also associated prediction Intervals (PIs) are important information for promoting the application of PV in practice, especially when grid connection continues to grow. While there are few studies focused on quantifying the uncertainty of forecasting PV power outputs, this paper developed a novel two-stage model to quantify the PIs of PV power outputs. In the first stage, three different neural networks, namely Generalized Regression Neural Network (GRNN), Extreme Learning Machine Neural Network (ELMNN) and Elman Neural Network (ElmanNN), were integrated using the Genetic Algorithms optimized Back Propagation (GA-BP) method to develop a Weight-Varying Combination Forecast Mode (WVCFM) model. The WVCFM model was applied to generate DPPs. In the second stage, a nonparametric kernel density estimation (NKDE) method was adopted to estimate the PIs regarding the statistical distribution of the errors of the DPPs derived in the first stage. The proposed method was tested using four types of PV output and weather data measured from a 15 kW grid-connected PV system. The cover percentage of prediction intervals (PICPs) were computed under the confidence level of 95%, 90%, 85% and 80%, respectively. The results imply that the two-stage model proposed in the paper outperforms conventional forecast methods in terms of prediction of short-term PV power outputs and associated uncertainties. [ABSTRACT FROM AUTHOR]

Highlights • A deep recurrent neural network (RNN) model is used for medium-term thermal load prediction. • The deep RNN model outperforms a simple multilayer perceptron for time series forecasting. • An optimization framework is proposed for sizing a thermal storage tank to meet thermal loads. • The optimization method can be used in selecting a thermal storage tank for use with a building. • The combined predictions and optimization allow for estimating performance of the tank. Abstract This paper evaluates the performance of deep recurrent neural networks in predicting heating demand for a commercial building over a medium-to-long term time horizon (⩾ 1 week), and proposes a modeling framework to demonstrate how these longer-term predictions can be used to aid design of a stratified thermal storage tank. The building sector contributes significantly to primary energy consumption in the US, and as such, there is a need to predict heating demand in buildings over longer time horizons, and to develop methods that can facilitate installation, planning and management of distributed generation and thermal storage to meet these heating demands. Key objectives of this paper are: (a) Investigate how a deep recurrent neural network model performs in predicting heating demand in campus buildings at University of Utah over multiple weeks, and (b) develop an optimization framework that which can provide definitive guidelines on sizing a stratified thermal storage tank without requiring high performance computing resources. The results showed that the predictions by the deep RNN are comparatively more accurate than those by a 3-layer MLP, and that these deep RNN predictions can adequately serve as proxy for future demand while considering sizing in the design of a complementary stratified thermal storage tank. [ABSTRACT FROM AUTHOR]

Microgrids are extensively integrated into electrical systems due to their many technical, economic, and environmental advantages. However, they encounter a challenge as they experience high-frequency fluctuations caused by the stochastic nature of renewable energy generation, electric loads, and the presence of Electric Vehicles (EVs). Therefore, various techniques, algorithms, and controllers have been introduced to ensure effective Load Frequency Control (LFC) and maintain a stable power system in microgrids. These methods aim to ensure that the system's frequency remains stable and within an acceptable range, especially when faced with changing load demands and other factors. This paper presents a novel enhanced control approach, Particle Swarm Optimization-Trained Artificial Neural Network (PSO-TANN), to optimize the load frequency model of a microgrid with vehicle-to-grid integration. The results are then compared under various scenarios, including renewable energy integration, EV charging and discharging dynamics, and varying load demands. The comparative analysis involves assessing the performance of the conventional Proportional–Integral–Derivative (PID) controller, the PSO-PID controller, and the newly proposed controlling technique. The suggested controller attains 99.904% efficiency with a negligible mean squared error of 1.1112 × 10−7, decreasing the integrated time absolute error to 1.0 × 10−4. It shows rapid response, precise targeting, and quick peak output ability, with marginal overshoot and undershoot, and a transient time of 28.5626 s, efficiently controlling microgrid frequency. Stability analysis validates the effectiveness of the proposed PSO-TANN controller in ensuring stability within the microgrid's LFC system during uncertainties and disturbances. This establishes resilience, diminishes settling time, and maintains reliable performance while controlling frequency. • An investigation into an islanded microgrid's Load Frequency Control (LFC) model framework has been conducted, comprising diesel generator, solar PV system, wind turbine, and EVs operating in vehicle-to-grid mode to supply power to electric loads, consequently impacting the frequency dynamics of the microgrid system. • A novel enhanced control approach, Particle Swarm Optimization-Trained Artificial Neural Network (PSO-TANN) has been implemented to optimize the load frequency model of a microgrid. • An extensive comparative analysis with other controllers, including PID and PSO-PID, is conducted across diverse scenarios involving renewable energy integration and EV charging and discharging dynamics under various load disturbances to reduce settling time, control system frequency, and ensure reliable performance. • The stability analysis validates the efficiency of the proposed PSO-TANN controller in ensuring reliability during uncertainties and disturbances, maintaining overall robustness and stability of the LFC system within the microgrid. [ABSTRACT FROM AUTHOR]

Electricity market liberalization and the absence of cost-efficient energy storage technologies have led to the transformation of state-owned electricity companies into complex electricity market entities, each having a different time horizon. Deregulation has intensified competition, giving rise to increased uncertainty caused by a multitude of interrelated exogenous factors, resulting in unexpected fluctuations in electricity prices. As a consequence, market participants encounter elevated risks and seek effective mitigation strategies. In this paper, the challenges described in the literature are addressed by studying price distribution histograms in the German and Finnish electricity markets. The objective is to identify normal price intervals that can serve as a foundation for an integrated Day-Ahead price forecasting methodology. A novel approach utilizing the Extreme Learning Machine in combination with Bootstrap intervals is proposed and applied to both markets. The findings demonstrate that Bootstrap intervals effectively capture normal prices, whereas extremely high prices typically align with the upper limits of Bootstrap intervals. Conversely, negative prices tend to fall outside the lower boundaries of the intervals. In order to assess the performance of the proposed methodology, a comparative analysis of its forecasting accuracy against the well-established Generalized AutoRegressive Conditional Heteroskedasticity and AutoRegressive Fractionally Integrated Moving Average models is conducted. In addition, both the computational efficiency and forecasting accuracy of the Extreme Learning Machine in comparison to the Artificial Neural Network are assessed. The results reveal the superior efficiency of the Extreme Learning Machine. The developed forecasting model could potentially assist market participants in making well-informed decisions and executing optimal bidding strategies in response to various scenarios before the Day-Ahead market closes. Notably, the proposed methodology transcends the limitations of fixed price thresholds and effectively addresses market nuances, including the occurrence of negative prices, thus offering a more comprehensive approach for electricity price forecasting. • Analysis of German & Finnish of Day-Ahead prices reveals new insights. • Proposed a new method of classifying market prices into unique classes. • ELM & Bootstrap forecast methodology with very good performance. • ELM & Bootstrap methodology outperforms traditional ones in accuracy & efficiency. [ABSTRACT FROM AUTHOR]

This paper presents hybrid control approaches for heating air supply in response to changes in demand by using the Fuzzy Inference System (FIS) and Artificial Neural Network (ANN) fitting models. Since early 2000’s, some advanced computing and statistical tools were introduced to replace conventional control models in improving control and energy efficiency. Among the tools, the FIS and ANN algorithms were used to define complex interactions between inputs and outputs, and were able to facilitate control models to predict or evaluate precise thermal performance. This paper introduces the FIS and ANN control schemes for simultaneously controlling the amount of supply air and its temperature. Input and output data derived from the FIS results generate and validate the ANN model, and both models are compared to the typical thermostat on/off baseline control to evaluate conditions of supply air for a heating season. The differences between the set-point and actual room temperature and their sums indicate control efficiency, and the heat gains into a room and their sums define the energy consumption level. This paper concludes that the simultaneous control of mass and temperature maintains the desired room temperature in a highly efficient manner. Sensitive controls may have a disadvantage in terms of energy consumption, but the ANN controller can minimize energy consumption in comparison with simple thermostat on/off controller. The results also confirm the effectiveness of simultaneous control of mass and temperature using an ANN algorithm corresponding to intermittent or unpredicted changes in thermal demands. [ABSTRACT FROM AUTHOR]

An accurate battery remaining useful life (RUL) estimation can facilitate the design of a reliable battery system as well as the safety and reliability of actual operation. A reasonable definition and an effective prediction algorithm are indispensable for the achievement of an accurate RUL estimation result. In this paper, the analysis of battery terminal voltage curves under different cycle numbers during charge process is utilized for RUL definition. Moreover, the relationship between RUL and charge curve is simulated by feed forward neural network (FFNN) for its simplicity and effectiveness. Considering the nonlinearity of lithium-ion charge curve, importance sampling (IS) is employed for FFNN input selection. Based on these results, an online approach using FFNN and IS is presented to estimate lithium-ion battery RUL in this paper. Experiments and numerical comparisons are conducted to validate the proposed method. The results show that the FFNN with IS is an accurate estimation method for actual operation. [ABSTRACT FROM AUTHOR]

Abstract: This paper reports an artificial neural networks (ANN) modelling programme for a light-duty diesel engine powered using blends of various biodiesel fuels with conventional fossil diesel. ANN was used here to predict nine different engine-out responses, namely carbon monoxide (CO), carbon dioxide (CO2), nitrogen monoxide (NO), unburned hydrocarbon (UHC), maximum pressure (P max), location of maximum pressure (CAD P max), maximum heat release rate (HRRmax), location of maximum HRR (CAD HRRmax) and cumulative HRR (CuHRR). Four pertinent engine operating parameters, engine speed, output torque, fuel mass flow rate and biodiesel fuel types and blends, were used as the input parameters for this modelling work. The feasibility of using ANN in predicting the relationships between these inputs and outputs were assessed. Simulated results were first validated against data from parallel engine test-bed study. Key effects of ANN “model” and “model parameter” such as type of transfer function, training algorithm and number of neurons, along with the methods of optimising the network settings were also presented in this paper. [Copyright &y& Elsevier]

Abstract: In this paper, a new multi-level genetic programming (MLGP) approach is introduced for forecasting transport energy demand (TED) in Iran. It is shown that the result obtained here has smaller error compared with the result obtained using neural network or fuzzy linear regression approach. The forecast uses historical energy data from 1968 to 2002 and it is based on three parameters; gross domestic product (GDP), population (POP), and the number of vehicles (VEH). The approach taken in this paper is based on genetic programming (GP) and the multi-level part of the name comes from the fact that we use GP in two different levels. At the first level, GP is used to obtain the time series model of the three parameters, GDP, POP, and VEH, and forecast those parameters for the time interval that their actual data are not available, and at the second level GP is used one more time to forecast TED based on available data for TED along with the data that are either available or predicted for the three parameters discussed earlier. Actual data from 1968 to 2002 are used for training and the data for years 2003–2005 are used to test the GP model. We have limited ourselves to these data ranges so that we could compare our results with the existing ones in the literature. The estimation GP for the model is formulated as a nonlinear optimization problem and it is solved numerically. [Copyright &y& Elsevier]

• Novel climate control framework for thermal comfort with minimum energy consumption. • Neural network well models ACMV energy with air temperature and operating frequencies. • Thermal comfort prediction improves building climate control and energy efficiency. • Proposed optimization algorithm exhibits 36.5% energy saving potential. • 25 °C is optimum air temperature for comfort-energy balance. Building air-conditioning and mechanical ventilation (ACMV) systems are responsible for significant energy consumption and yet, dissatisfaction with the thermal environment is prevalent among the occupants, revealing a widespread disparity between energy-efficiency and indoor thermal-comfort in buildings. This paper presents an indoor-climate control framework that bridges this gap between energy and comfort. The framework comprises two main components: a thermal-comfort prediction model, and an optimization algorithm termed as the optimal air temperature (OAT) algorithm; they collectively act as an intelligent mediator between the occupant and the ACMV system. Firstly, the ACMV energy consumption is modelled as a function of air temperature, and three operating frequencies of cooling components using a feedforward neural network. Secondly, the thermal-comfort prediction model predicts the thermal state index (TSI: Cool-Discomfort/Comfort/Warm-Discomfort). Thirdly, depending on the predicted TSI, the OAT algorithm locates the optimal operating state such that Comfort state is achieved using the minimum ACMV energy consumption. Proposed framework exhibits an energy saving potential of 36.5%. It is found that 25 °C is the ideal air temperature for desired comfort with minimum energy expense in the tropical buildings. Additionally, six different TSI predictive models including two general and four personal comfort models are implemented to validate the framework. The study is substantiated with extensive real human experiments in controlled thermal environment. The proposed method is scalable for its applicability with any comfort-prediction model, and adaptive for its data-driven architecture. It exhibits the potential to achieve both occupant-comfort and energy-saving through integration with the Internet-of-Things for realizing comfort-energy balanced buildings. [ABSTRACT FROM AUTHOR]

• A novel vector field-based SVR for building energy consumption prediction (ECP) is proposed. • Nonlinearity of the ECP models is transformed into linearity via vector field-based SVR. • Two high precision, generalization capability, and robustness energy consumption prediction models are established. • A real case study demonstrates the superiority of this study compared with previous studies. Building energy consumption prediction plays an irreplaceable role in energy planning, management, and conservation. Data-driven approaches, such as artificial neural networks, support vector regression, gradient boosting regression and extreme learning machine are the most advanced methods for building energy prediction. However, owing to the high nonlinearity between inputs and outputs of building energy consumption prediction models, the aforementioned approaches require improvement with regard to the prediction accuracy, robustness, and generalization ability. To counter these shortcomings, a novel vector field-based support vector regression method is proposed in this paper. Through multi-distortions in the sample data space or high-dimensional feature space mapped by a vector field, the optimal feature space is found, in which the high nonlinearity between inputs and outputs is approximated by linearity. Hence, the proposed method ensures a high accuracy, a generalization ability, and robustness for building energy consumption prediction. A large office building in a coastal town of China is used for a case study, and its summer hourly cooling load data are used as energy consumption data. The results indicate that the proposed method achieves better performance than commonly used methods with regard to the accuracy, robustness, and generalization ability. [ABSTRACT FROM AUTHOR]

• Energy prediction model has been combined with estimated occupancy profile. • Blind system identification models give accurate estimation of occupancy. • Extra input of occupancy can improve the accuracy of electricity prediction models. • Best performance is obtained with ensemble model in all neural network structures. Occupancy behaviour plays an important role in energy consumption in buildings. Currently, the shallow understanding of occupancy has led to a considerable performance gap between predicted and measured energy use. This paper presents an approach to estimate the occupancy based on blind system identification (BSI), and a prediction model of electricity consumption by an air-conditioning system is developed and reported based on an artificial neural network with the BSI estimation of the number of occupants as an input. This starts from the identification of indoor C O 2 dynamics derived from the mass-conservation law and venting levels. The unknown parameters, including the occupancy and model parameters, are estimated by using a frequentist maximum-likelihood algorithm and Bayesian estimation. The second phase is to establish the prediction model of the electricity consumption of the air-conditioning system by using a feed-forward neural network (FFNN) and extreme learning machine (ELM), as well as ensemble models. To analyse some aspects of the benchmark test for identifying the effect of structure parameters and input-selection alternatives, three studies are conducted on (1) the effect of predictor selection based on principal component analysis, (2) the effect of the estimated occupancy as the supplementary input, and (3) the effect of the neural network ensemble. The result shows that the occupancy number, as the input, is able to improve the accuracy in predicting energy consumption using a neural-network model. [ABSTRACT FROM AUTHOR]

Highlights • The spatio-temporal feature is proposed, instead of traditional feature, time series. • Convolutional network is used to predict wind power based on spatio-temporal feature. • Much higher accuracy is achieved within much less training time than existing works. Abstract Wind power prediction is of vital importance in wind power utilization. There have been a lot of researches based on the time series of the wind power or speed. But in fact, these time series cannot express the temporal and spatial changes of wind, which fundamentally hinders the advance of wind power prediction. In this paper, a new kind of feature that can describe the process of temporal and spatial variation is proposed, namely, spatio-temporal feature. We first map the data collected at each moment from the wind turbines to the plane to form the state map, namely, the scene, according to the relative positions. The scene time series over a period of time is a multi-channel image, i.e. the spatio-temporal feature. Based on the spatio-temporal features, the deep convolutional network is applied to predict the wind power, achieving a far better accuracy than the existing methods. Compared with the state-of-the-art methods, the mean-square error in our method is reduced by 49.83%, and the average time cost for training models can be shortened by a factor of more than 150. [ABSTRACT FROM AUTHOR]

Highlights • Set multi-objective for electrical load forecasting. • Propose a hybrid model contains based on the multi-objective optimization. • Use data processing technique to pre-process original data. Abstract To ensure the safe operation of electrical power systems, short-term load forecasting (STLF) plays a significant role. With the development of artificial neural network (ANN), many forecasting models based on ANN are proposed to enhance the forecasting accuracy. However, forecasting stability is also an important aspect when considering a forecasting model. Both forecasting accuracy and stability are affected heavily by the random initial values of weights and thresholds of ANN. Thus, in this paper, a new hybrid model based on the modified generalized regression neural network (GRNN) is proposed for short-term load forecasting (STLF). Meanwhile, a non-dominated sorting-based multi-objective cuckoo search algorithm (NSMOCS) is proposed to realize accurate and stable forecasting simultaneously. To utilize the similarities and reduce interference existing in the original data, some data pre-processing techniques are also incorporated. With half-hourly load data from five states in Australia, experimental results clearly show that the proposed hybrid model could obtain more accurate and stable forecasting results, compared with the comparison models. [ABSTRACT FROM AUTHOR]

Highlights • A probabilistic load forecasting model for buildings is developed. • Weather forecasting uncertainty is quantified and used for load forecasting. • Uncertain peak load is modeled considering its magnitude and occurrence. • A hybrid method is proposed to decompose the original load dataset. • Case studies are conducted to compare different means of using weather forecasts. Abstract This paper presents a new probabilistic building load forecasting model considering uncertainties in weather forecasts and abnormal peak load. Two basis function components are developed, including the probabilistic normal load forecasting and the probabilistic uncertain peak load (or abnormal peak load) forecasting. The probabilistic normal load forecasting model is built using the artificial neural network (ANN) and the probabilistic temperature forecasts. The probabilistic abnormal peak load forecasting model consists of two models quantifying the probabilistic occurrence and magnitude of the peak abnormal differential load respectively. The test results show that the ANN deterministic load forecasting model can achieve satisfactory performance with the average mean absolute percentage error (MAPE) of 5.0%. The probabilistic occurrence model can forecast the occurrence frequency of the peak abnormal differential load with the satisfactory agreement, and the probabilistic magnitude model can well forecast the magnitudes of the peak abnormal differential load with the Kolmogorov-Smirnov error of 0.09. Real-time application case studies are conducted by different means of using the probabilistic weather forecasts. The results show that the probabilistic normal load forecasts have satisfactory accuracies and the load forecasts based on the one-day-ahead probabilistic weather forecasts are the best. [ABSTRACT FROM AUTHOR]

Highlights • An optimisation strategy that controls both the energy supply and the energy demand. • Artificial neural networks predict the energy demand and renewable energy supply. • An error management stage adjusts the solution to react to prediction errors. • Resulted in a 52.92% increase in profit while reducing CO 2 emissions by 3.75%. Abstract Decentralisation of energy generation and distribution to local districts or microgrids is viewed as an important strategy to increase energy efficiency, incorporate more small-scale renewable sources and reduce greenhouse gas emissions. To achieve these goals, an intelligent, context-aware, adaptive energy management platform is required. This paper will demonstrate two district energy management optimisation strategies; one that optimises district heat generation from a multi-vector energy centre and a second that directly controls building demand via the heating set point temperature in addition to the heat generation. Several Artificial Neural Networks are used to predict variables such as building demand, solar photovoltaic generation, and indoor temperature. These predictions are utilised within a Genetic Algorithm to determine the optimal operating schedules of the heat generation equipment, thermal storage, and the heating set point temperature. Optimising the generation of heat for the district led to a 44.88% increase in profit compared to a rule-based, priority order baseline strategy. An additional 8.04% increase in profit was achieved when the optimisation could also directly control a proportion of building demand. These results demonstrates the potential gain when energy can be managed in a more holistic manner considering multiple energy vectors as well as both supply and demand. [ABSTRACT FROM AUTHOR]

Highlights • A closed-loop deep-learning method for PQD classification is proposed. • A special unit construction is designed to facilitate the identification of PQDs. • The deep CNN is applied to realize an automatic feature extraction and selection. • The proposed method has advantages in accuracy, efficiency and noise immunity. Abstract With the integration of multiple energy systems, there are more and more deterioration risks of power quality in different energy production, transformation, delivery and consumption stages. Automatic classification of power quality disturbances is the foundation to deal with power quality problem. From the traditional point of view, the identification process of power quality disturbances should be divided into three independent stages: signal analysis, feature selection and classification. However, there are some inherent defects in signal analysis and the procedure of manual feature selection is tedious and imprecise, leading to a low classification accuracy of multiple disturbances and a poor noise immunity. This paper proposes a novel full closed-loop approach to detect and classify power quality disturbances based on a deep convolutional neural network. Considering the characteristics of power quality disturbances problem, a unit construction which consists of 1-D convolutional, pooling, and batch-normalization layers is designed to capture multi-scale features and reduce overfitting. In the proposed deep convolutional neural network, multiple units are stacked to extract features from massive disturbance samples automatically. Comparisons with other state-of-the-art deep neural networks and traditional methods proves that the proposed method can overcome defects of traditional signal process and artificial feature selection. Considering microgrid is an important development form of multi-energy system and an essential part of smart grid, a typical simulation system is constructed to analyze the causes of power quality problems in microgrid and the field data from a multi-microgrid system are used to further prove the validity of the proposed method. [ABSTRACT FROM AUTHOR]

Highlights • Estimation of U -value by means of artificial neural network. • The model developed does not need a data post-processing. • Representative results for all the building periods considered. Abstract Most of the existing building stock has a deficient energy behaviour. The thermal transmittance of façades is among those aspects which most affect this situation. In this paper, the calculation procedure with correction for storage effects from ISO 9869-1 was applied to the thermometric method to determine the U -value. Due to the need for determining the number and type of layers that compose the wall to apply the calculation, a multilayer perceptron has been developed to estimate the U -value. From the different model configurations suggested, the most adequate architecture was the one with 14 nodes in the hidden layer without making transformations in the input variables. Valid results have been obtained by the multilayer perceptron for the case studies analysed from different building periods, with deviations lower than 20% between the measured value and the expected one, varying the test duration according to the thermal resistance of the wall and the temperature variations. Furthermore, it is not necessary to carry out a data post-processing for the model, so this fact simplifies and hastens the calculation procedure. [ABSTRACT FROM AUTHOR]

Highlights • LASSO-QRNN is proposed for electricity consumption probability density forecasting. • LASSO regression is used to select the important variables from the external factors. • Two cases of Guangdong province and California show the performance of proposed method. Abstract The electricity consumption forecasting is a challenging task, because the predictive accuracy is easily affected by multiple external factors, such as society, economics, environment, as well as the renewable energy, including hydro power, wind power and solar power. Particularly, in the smart grid with large amount of data, how to extract valuable information of those external factors timely is the key to the success of electricity consumption forecasting. A method of probability density forecasting based on Least Absolute Shrinkage and Selection Operator-Quantile Regression Neural Network (LASSO-QRNN) is proposed in this paper. First, important features are extracted from external factors affecting the electricity consumption forecasting by LASSO regression. Then, the LASSO-QRNN model is constructed to predict annual electricity consumption. The results of electricity consumption forecasting under different quantiles in the next several years are evaluated. Besides, we introduce kernel density estimation into our LASSO-QRNN model, which can give a probability distribution instead of a single-valued prediction. The prediction accuracy is evaluated through the empirical analyses from the Guangdong province dataset in China and the California dataset in the United States. The simulation results demonstrate that the proposed method provides better performance for electricity consumption forecasting, in comparison with existing quantile regression neural network (QRNN), back-propagation of errors neural network (BP), radial basis function neural network (RBF), quantile regression (QR) and nonlinear quantile regression (NLQR). LASSO-QRNN can not only better learn the high-dimensional data in electricity consumption forecasting, but also provide more precise results. [ABSTRACT FROM AUTHOR]

Abstract: This paper demonstrates the use of artificial neural networks virtual sensors in emissions prediction and control for a gasoline engine. Tailpipe emissions and engine parameters were first measured experimentally to form a comprehensive database for network training and testing. Individual predictive models were constructed using the optimization layer-by-layer neural network. Simulation results demonstrated that the networks, as virtual sensors, can accurately predict the engine parameters and emissions quantitatively and qualitatively with RMS errors below 9%. The second part of this paper then presents a virtual sensor control model which is the combination of the two individual emissions and engine predictive models developed previously. The main objective of this part is to control the exhaust emissions within the desired limits by predicting optimum engine parameters with the use of artificial neural network virtual sensors. Results showed that the emissions levels were successfully controlled within the defined limits, with maximum tolerance of 6%. This first part of this paper demonstrated that with the use of artificial neural network virtual sensors, emissions and engine parameters can be accurately predicted. Hence with accurate virtual sensors, emissions were then controlled within the desired limits by optimizing the engine parameters. This proposed work demonstrated a viable and accurate methodology in emissions predictive and control. By applying virtual sensor models, the need additional, cumbersome and costly measuring and monitoring devices can be eliminated. [Copyright &y& Elsevier]

Abstract: This paper presents an adaptive network-based fuzzy inference system-stochastic frontier analysis (ANFIS-SFA) approach for long-term natural gas (NG) consumption prediction and analysis of the behavior of NG consumption. The proposed models consist of input variables of Gross Domestic Product (GDP) and population (POP). Six distinct models based on different inputs are defined. All of trained ANFIS are then compared with respect to mean absolute percentage error (MAPE). To meet the best performance of the intelligent based approaches, data are pre-processed (scaled) and finally the outputs are post-processed (returned to its original scale). To show the applicability and superiority of the integrated ANFIS-SFA approach, gas consumption in four Middle Eastern countries i.e. Bahrain, Saudi Arabia, Syria, and United Arab Emirates is forecasted and analyzed based on the data of the time period 1980–2007. With the aid of autoregressive model, GDP and population are projected for the period 2008–2015. These projected data are used as the input of ANFIS model to predict the gas consumption in the selected countries for 2008–2015. SFA is then used to examine the behavior of gas consumption in the past and also to make insights for the forthcoming years. The ANFIS-SFA approach is capable of dealing with complexity, uncertainty, and randomness as well as several other unique features discussed in this paper. [Copyright &y& Elsevier]

Abstract: Wind energy evaluation is an important goal in the conversion of energy systems to more environmentally friendly solutions. In this paper, we present a novel approach to wind speed spatial estimation on the isle of Sicily (Italy): an incremental self-organizing neural network (Generalized Mapping Regressor – GMR) is coupled with exploratory data analysis techniques in order to obtain a map of the spatial distribution of the average wind speed over the entire region. First, the topographic surface of the island was modelled using two different neural techniques and by exploiting the information extracted from a digital elevation model of the region. Then, GMR was used for automatic modelling of the terrain roughness. Afterwards, a statistical analysis of the wind data allowed for the estimation of the parameters of the Weibull wind probability distribution function. In the last sections of the paper, the expected values of the Weibull distributions were regionalized using the GMR neural network. [Copyright &y& Elsevier]

• A comprehensive literature review of meta-heuristic optimization algorithms for wind power forecasting is conducted. • A well-designed taxonomy of meta-heuristic algorithms for optimizing wind power forecasting model parameters is developed. • A deeply comprehensive and scientific multiple error evaluation metric is proposed for wind power error analysis. • Further research topics for future works in wind power forecasting are presented. The integration of large-scale wind power introduces issues in modern power systems operations due to its strong randomness and volatility. These issues can be resolved via wind power forecasting that can provide comprehensive future information about wind power uncertainties. This paper presents a timely and comprehensive review of meta-heuristic algorithms in the framework of wind power forecasting. The framework is based on the auxiliary layer, forecasting base layer, and core layer. The auxiliary layer, such as the data-decomposition layer, decomposes the wind power time series into many relatively stationary subseries, and uses prediction models, including artificial neural networks (ANNs) and machine learning (ML). The core layer is based on meta-heuristic algorithms, which include evolutionary-based algorithms, physics-based algorithms, human-based algorithms, swarm-based algorithms, hybrid algorithms, and multi-objective optimization algorithms. These algorithms aim to search for the optimal solutions under constraints, which is highly significant for optimizing the key parameters of the prediction models. Besides, multiple error evaluation metrics, e.g., deterministic, uncertainty, and testing methods used in the field of wind power prediction are described. A quantitative analysis focusing on their advantages, disadvantages, forecasting accuracy, and computational costs are also provided. Finally, a few open research issues and trends related to the topic are discussed, which can contribute to improving the understanding of each wind power forecasting method. In general, this review paper provides valuable insights to wind power engineers. [ABSTRACT FROM AUTHOR]

The Tristar aircraft, operated by the Royal Air Force, fly many thousands of hours per year in the transport and air-to-air refuelling roles. A large amount of engine data is recorded for each of the Rolls-Royce RB211-524B4 engines: it is used to aid the maintenance process. Data are also generated during test-bed engine ground-runs after repair and overhaul. In order to use recorded engine data more effectively, this paper assesses the feasibility of a pro-active engine diagnostic-tool using artificial neural networks (ANNs). Engine-health monitoring is described and the theory behind an ANN is described. An engine diagnostic structure is proposed using several ANNs. The top level distinguishes between single-component faults (SCFs) and double-component faults (DCFs). The middle-level class includes components, or component pairs, which are faulty. The bottom level estimates the values of the engine-independent parameters, for each engine component, based on a set of engine data using dependent parameters. The DCF results presented in this paper illustrate the potential for ANNs as diagnostic tools. However, there are also a number of features of ANN applications that are user-defined: ANN designs; the number of training epochs used; the training function employed; the method of performance assessment; and the degree of deterioration for each engine-component''s performance parameter. [Copyright &y& Elsevier]

• We develop shallow and deep neural networks for the Short-Term Load Forecasting problem. • Different neural networks architectures are tested, including uni- and bi-directional structures. • Global climate models' information is used as input of the neural networks. • We present a real study case of time series forecasting for the Brazilian power system. • Relevant results are presented and systematically compared using Diebold-Mariano test. This paper focuses on the development of shallow and deep neural networks in the form of multi-layer perceptron, long-short term memory, and gated recurrent unit to model the short-term load forecasting problem. Different model architectures are tested, and global climate model information is used as input to generate more accurate forecasts. A real study case is presented for the Brazilian interconnected power system and the results generated are compared with the forecasts from the Brazilian Independent System Operator model. In general terms, results show that the bidirectional versions of long-short term memory and gated recurrent unit produce better and more reliable predictions than the other models. From the obtained results, the recurrent neural networks reach Nash-Sutcliffe values up to 0.98, and mean absolute percentile error values of 1.18%, superior than the results obtained by the Independent System Operator models (0.94 and 2.01% respectively). The better performance of the neural network models is confirmed under the Diebold-Mariano pairwise comparison test. [ABSTRACT FROM AUTHOR]

Highlights • Developing building cooling forecasting model for energy saving of HVAC system. • Forecasting step-ahead building cooling load by time series prediction. • Forecasting step-ahead building cooling load by Fourier’s law-based analysis. • Improving performance of building cooling load forecasting by ensemble technique. Abstract Short-term load prediction, which forecasts a building’s thermal load with a lead time ranging from seconds to a few days, is essential for not only monitoring and controlling the system operation, but also on-line scheduling. Dynamic cooling load forecasting, which belongs to short-term load prediction, is both meaningful for monitoring the system or fuzzy on-line scheduling and crucial for solving the time-lag problem to meet the heating, ventilation and air-conditioning system’s time-varying cooling loads. Numerous studies have been carried out to develop dynamic load-forecasting models, and great achievements have been made. However, limitations in their applicability persist because most previous models are calendar- and time-based data-driven models that may fail when unexpected issues occur or special schedules are adopted. What’s more, the inputs that were selected passively from the source data pools at hand rather than via active exploration may be insufficient and impair the accuracy of forecasting models. This paper proposes a novel dynamic forecasting model for building cooling loads that combines an artificial neural network with an ensemble approach. Based on physical principles other than the available data source, the inputs are explored actively and are independent from both calendar and time indicators, which make the forecasting model being capable of dealing with irregular occasions and unexpected schedules with high accuracy. A benchmark is proposed that uses the current load Q (t) as a forecasted cooling load Q ^ (t + i) and gives the minimum accuracy requirement for a dynamic forecasting model. The benchmark not only can be used to evaluate dynamic forecasting models that are validated by various case studies, but also ensures that the proposed forecasting model can be applied immediately to heating, ventilation and air-conditioning systems to tackle the time-lag problem. [ABSTRACT FROM AUTHOR]

Highlights • The complexity of electric market models limits their usefulness in risk analysis. • A reduced-form modeling approach is developed using neural networks. • The reduced-form model enables the use of modern, simulation-based risk analysis. • Applications to the analysis of the cash flow risks of generators are explored. Abstract The experience of the last fifteen years has illustrated dramatically the emergence of new risks facing power market investors. The volatility of commodity prices, the strategic behavior of competing firms, and regulatory uncertainty all contribute to a challenging investment and operating environment. Traditionally, utilities and power-market investors have used large-scale optimizing production-cost models to analyze the cash flows of power generators. The complexity of these models, particularly when applied on a regional or national scale, is such that computational costs often prohibit extensive analysis of commodity, regulatory, and structural risks. This article demonstrates how a reduced-form modeling approach utilizing neural networks can be used to increase greatly the ability of modelers to use modern simulation-based risk analysis techniques. In particular, several applications relevant to evaluating the cash flow risks of generators, with applications to hedging, are presented. Central to the contributions of this paper is our reduction of complex optimizing models to spreadsheet form, reducing not only their computational complexity, but also their practical user complexity. [ABSTRACT FROM AUTHOR]

This paper presents a novel method for the development of multi-step wind forecasting models based on functional network (FN), a modern intelligent paradigm. The basis of FN development is the integration of functional theory with neural networks to produce problem-driven network topologies and optimal neural functions with diversified structures as opposed to conventional neural networks. These advantages of functional networks result in optimum models for accurate wind speed and power forecasting. In this research work, FN forecasting engine is developed using three state-of-the-art multi-step forecasting mechanisms, namely, recursive, direct and hybrid DirRec scheme. A detailed analysis of the developed forecast models is carried out using a real-world case study and notable improvement in forecast accuracy is recorded in terms of standard performance indices. Among the three multi-step schemes, hybrid DirRec gives the best forecast accuracy. The results obtained from a comparative analysis against a benchmark model as well as a classical neural network model validate the efficacy of the FN model. Hence the proposed forecasting schemes can be of immense utility for wind power system operators for devising cost-effective energy management and dispatch strategies by accurately forecasting wind power for long forecast horizons. [ABSTRACT FROM AUTHOR]

Energy demand prediction of building heating is conducive to optimal control, fault detection and diagnosis and building intelligentization. In this study, energy demand prediction models are developed through machine learning methods, including extreme learning machine, multiple linear regression, support vector regression and backpropagation neural network. Seven different meteorological parameters, operating parameters, time and indoor temperature parameters are used as feature variables of the model. Correlation analysis method is utilized to optimize the feature sets. Moreover, this paper proposes a strategy for obtaining the thermal response time of building, which is used as the time ahead of prediction models. The prediction performances of extreme learning machine models with various hidden layer nodes are analyzed and contrasted. Actual data of building heating using a ground source heat pump system are collected and used to test the performances of the models. Results show that the thermal response time of the building is approximately 40 min. Four feature sets are obtained, and the performances of the models with feature set 4 are better. For different machine learning methods, the performances of extreme learning machine models are better than others. In addition, the optimal number of hidden layer nodes is 11 for the extreme learning machine model with feature set 4. [ABSTRACT FROM AUTHOR]

Many studies have proven that the building sector can significantly benefit from replacing the current practice rule-based controllers (RBC) by more advanced control strategies like model predictive control (MPC). However, the optimization-based control algorithms, like MPC, impose increasing hardware and software requirements, together with more complicated error handling capabilities required from the commissioning staff. In recent years, several studies introduced promising remedy for these problems by using machine learning algorithms. The idea is based on devising simplified control laws learned from MPC. The main advantage of the proposed methods stems from their easy implementation even on low-level hardware. However, most of the reported studies were dealing only with problems with a limited complexity of the parametric space, and devising laws only for a single control variable, which inevitably limits their applicability to more complex building control problems. In this paper, we introduce a versatile framework for synthesis of simple, yet well-performing control strategies that mimic the behavior of optimization-based controllers, also for large scale multiple-input-multiple-output (MIMO) control problems which are common in the building sector. The approach employs multivariate regression and dimensionality reduction algorithms. Particularly, deep time delay neural networks (TDNN) and regression trees (RT) are used to derive the dependency of multiple real-valued control inputs on parameters. The complexity of the problem, as well as implementation cost, are further reduced by selecting the most significant features from the set of parameters. This reduction is based on straightforward manual selection, principal component analysis (PCA) and dynamic analysis of the building model. The approach is demonstrated on a case study employing temperature control in a six-zone building, described by a linear model with 286 states and 42 disturbances, resulting in an MPC problem with more than thousand of parameters. The results show that simplified control laws retain most of the performance of the complex MPC, while significantly decreasing the complexity and implementation cost. [ABSTRACT FROM AUTHOR]

Due to the explosion in restructuring of power markets within a deregulated economy, competitive power market needs to minimize their required generation reserve gaps. Efficient load forecasting for future demands can minimize the gap which will help in economic power generation, power operations, power construction planning and power distribution. Nowadays, neural networks are widely used for solving load forecasting problem due to its non-linear characteristics. Consequently, neural network is successfully combined with optimization techniques for finding optimal network parameters in order to reduce the forecasting error. In this paper, firstly a novel evolutionary algorithm based on follow the leader concept is developed and thereafter its performance is validated by COmparing Continuous Optimizers experimental framework on the set of 24 Black-Box Optimization Benchmarking functions with 12 state-of-art algorithms in 2-D, 3-D, 5-D, 10-D, and 20-D. The proposed algorithm outperformed all state-of-art algorithms in 20-D and ranked second in other dimensions. Further, the proposed algorithm is integrated with neural network for the proper tuning of network parameters to solve the real world problem of short term load forecasting. Through experiments on three real-world electricity load data sets namely New Pool England, New South Wales and Electric Reliability Council of Texas, we compared our proposed hybrid approach to baseline approaches and demonstrated its effectiveness in terms of predictive accuracy measures. [ABSTRACT FROM AUTHOR]

Building performance simulations (BPS) are used to test different designs and systems with the intention of reducing building costs and energy demand while ensuring a comfortable indoor climate. Unfortunately, software for BPS is computationally intensive. This makes it impractical to run thousands of simulations for sensitivity analysis and optimization. Worse yet, millions of simulations may be necessary for a thorough exploration of the high-dimensional design space formed by the many design parameters. This computational issue may be overcome by the creation of fast metamodels. In this paper, we aim to find suitable metamodeling techniques for diverse outputs from BPS. We consider five indicators of building performance and eight test problems for the comparison six popular metamodeling techniques – linear regression with ordinary least squares (OLS), random forest (RF), support vector regression (SVR), multivariate adaptive regression splines, Gaussian process regression (GPR), and neural network (NN). The methods are compared with respect to accuracy, efficiency, ease-of-use, robustness, and interpretability. To conduct a fair and in-depth comparison, a methodological approach is pursued using exhaustive grid searches for model selection assisted by sensitivity analysis. The comparison shows that GPR produces the most accurate metamodels, followed by NN and MARS. GPR is robust and easy to implement but becomes inefficient for large training sets compared to NN and MARS. A coefficient of determination, R 2 , larger than 0.9 have been obtained for the BPS outputs using between 128 and 1024 training points. In contrast, accurate metamodels with R 2 values larger than 0.99 can be achieved for all eight test problems using only 32–256 training points. [ABSTRACT FROM AUTHOR]

Motivated by the increasing integration among electricity markets, in this paper we propose two different methods to incorporate market integration in electricity price forecasting and to improve the predictive performance. First, we propose a deep neural network that considers features from connected markets to improve the predictive accuracy in a local market. To measure the importance of these features, we propose a novel feature selection algorithm that, by using Bayesian optimization and functional analysis of variance, evaluates the effect of the features on the algorithm performance. In addition, using market integration, we propose a second model that, by simultaneously predicting prices from two markets, improves the forecasting accuracy even further. As a case study, we consider the electricity market in Belgium and the improvements in forecasting accuracy when using various French electricity features. We show that the two proposed models lead to improvements that are statistically significant. Particularly, due to market integration, the predictive accuracy is improved from 15.7% to 12.5% sMAPE (symmetric mean absolute percentage error). In addition, we show that the proposed feature selection algorithm is able to perform a correct assessment, i.e. to discard the irrelevant features. [ABSTRACT FROM AUTHOR]

This paper explores the optimization scheme of carbon trading in China based on a novel energy-saving and emission-reduction (ESER) system with carbon price constraints. With the aid of nonlinear dynamics theory, the dynamics behavior of the novel system is discussed. Genetic algorithm and back propagation neural network is used to identify the quantitative coefficients according to the statistical data of the second period in European Union (EU). Taking the actual situation in EU for instance, the variables which are sensitive to carbon trading are detailedly researched. Enlightened by the EU’s experience, an optimal road of China ’ s carbon trading is put forward. The results show that carbon emissions could be controlled by carbon trading. The investment to carbon trading hampers economic growth in the near future, and ESER technical progress is negatively correlated with carbon trading in the long run. Demand and supply relationship is closely related to carbon price, both are the important issues in carbon trading system. Excessive government control and extortionate carbon price will deliver the opposite effect and even fatal influence on carbon trading system. [ABSTRACT FROM AUTHOR]

In the competitive power industry environment, electricity price forecasting is a fundamental task when market participants decide upon bidding strategies. This has led researchers in the last years to intensely search for accurate forecasting methods, contributing to better risk assessment, with significant financial repercussions. This paper presents a hybrid method that combines similar and recent day-based selection, correlation and wavelet analysis in a pre-processing stage. Afterwards a feedforward neural network is used alongside Bat and Scaled Conjugate Gradient Algorithms to improve the traditional neural network learning capability. Another feature is the method’s capacity to fine-tune neural network architecture and wavelet decomposition, for which there is no optimal paradigm. Numerical testing was applied in a day-ahead framework to historical data pertaining to Spanish and Pennsylvania-New Jersey-Maryland (PJM) electricity markets, revealing positive forecasting results in comparison with other state-of-the-art methods. [ABSTRACT FROM AUTHOR]

Various models based on measure-correlate-predict (MCP) methods have been used to estimate the long-term wind turbine power output (WTPO) at target sites for which only short-term meteorological data are available. The MCP models used to date share the postulate that the influence of air density variation is of little importance, assume the standard value of 1.225 kg m −3 and only consider wind turbines (WTs) with blade pitch control. A performance assessment is undertaken in this paper of the models used to date and of newly proposed models. Our models incorporate air density in the MCP model as an additional covariable in long-term WTPO estimation and consider both WTs with blade pitch control and stall-regulated WTs. The advantages of including this covariable are assessed using different functional forms and different machine learning algorithms for their implementation (Artificial Neural Network, Support Vector Machine for regression and Random Forest). The models and the regression techniques used in them were applied to the mean hourly wind speeds and directions and air densities recorded in 2014 at ten weather stations in the Canary Archipelago (Spain). Several conclusions were drawn from the results, including most notably: (a) to clearly show the notable effect of air density variability when estimating WTPOs, it is important to consider the functional ways in which the features air density and wind speed and direction intervene, (b) of the five MCP models under comparison, the one that separately estimates wind speeds and air densities to later predict the WTPOs always provided the best mean absolute error, mean absolute relative error and coefficient of determination metrics, independently of the target station and type of WT under consideration, (c) the models which used Support Vector Machines (SVMs) for regression or random forests (RFs) always provided better metrics than those that used artificial neural networks, with the differences being statistically significant (5% significance) for most of the cases assessed, (d) no statistically significant differences were found between the SVM- and RF-based models. [ABSTRACT FROM AUTHOR]

Assessment of solar potential over a location of interest is an important step towards the successful planning of renewable energy projects. However, solar data are not available for every point of interest due to the absence of meteorological stations and sophisticated solar sensors, so solar radiation has to be estimated using models. This paper presents a hybrid technique to improve the performance of a widely used modeling technique i.e. artificial neural network (ANN). Four different architectures of ANN, namely: multilayer perceptron (MLP), Adaptive neuro-fuzzy inference system (ANFIS), Nonlinear autoregressive recurrent exogenous neural network (NARX), and generalized regression neural networks (GRNN), are used in this study. A wavelet multiresolution analysis is applied to decompose the complex meteorological signals into relatively simple parts, wavelet sub-series, using discrete wavelet transformation (DWT) algorithm. The wavelet sub-series are modeled by the ANN models and reconstructed to estimate the original signal. Hence, enhancing the learning process of these models. Four meteorological parameters, namely: temperature (T), relative humidity (RH), wind speed (WS), and sunshine duration (SSD), are used to mode the global horizontal irradiation (GHI) over Abu Dhabi, the United Arab Emirates. The proposed approach is compared to standalone ANN models and validated using well-known statistical validation metrics including coefficient of determination (R 2 ), root mean square error (RMSE), mean bias error (MBE), mean absolute percentage error (MAPE), and t -statistics. In addition, wavelet cross spectrum (WCS) is used as a visual indicator of the model performance in time, frequency, and phase domains. The results show that using the proposed strategy considerably improves the modeling performance of the ANN with a maximum improvement of 6.84% in R 2 for MLP. In addition, minimum RMSE of 2.78% is observed for GRNN. [ABSTRACT FROM AUTHOR]

• Methodology and model for creating AI expert learning via AR and humans. • Positive climate impact when some services can be carried out remotely over AR calls. • Energy savings and CO₂ emissions reduction have been estimated. • Supporting the development of smart industries and autonomous operations. • Economic benefits as for health improvement and increased business productivity. Initially, Artificial Intelligence (AI) focused on diagnostics during the 70s and 80s. Unfortunately, it did not gain trust and few industries embraced it, mostly due to the extensive manual programming effort that AI required for interpreting data and act. In addition, the computer capacity, for handling the amounts of data necessary to train AI, was lacking the disc dimensions we are used to today, which made it go slowly. Not until the 2000 s confidence in AI was established in parallel with the introduction of new tools that was paving the way for PLS, PCA, ANN and soft sensors. Year 2011, IBM Watson (an AI application) was developed and won over the jeopardy champion. Today's machine learning (ML) such as "deep learning" and artificial neural networks (ANN) have created interesting use cases. AI has therefore regained confidence and industries are beginning to embrace where they see appropriate uses. Simultaneously, Internet of Things (IoT) tools have been introduced and made it possible to develop new capabilities such as virtual reality (VR), augmented reality (AR), mixed reality (MR) and extended reality (XR). These technologies are maturing and could be used in several application areas for the industries and form part of their digitalization journey. Furthermore, it is not only the industries that could benefit from introducing these technologies. Studies also show several areas and use cases where augmented reality has a positive impact, such as on students' learning ability. Yet few teachers know or use this technology. This paper evaluates and analyze AR, remote assistance tool for industrial purposes. The potential of the tool is discussed for frequent maintenance cases in the mining industry. Further on, if we look into the future, it is not surprising if we will be able to see that today's concepts of reality tools have evolved to become smarter by being trained by multimedia recognition and from people who have thus created an AI expert. Where the AI expert will support customers and be able to solve simple errors but also those that occur rarely and thus be a natural part of the solution for future completely autonomous processes for the industry. The article demonstrates a framework for creating smarter tools by combining AR, ML and AI and forms part of the basis creating the smarter industry of the future. Natural Language Processing (NLP) toolbox has been utilized to train and test an AI expert to give suitable resolutions to a specific maintenance request. The motivation for AR is the possible energy savings and reduction of CO₂ emissions in the maintenance field for all business trips that can be avoided. At the same time saving money for the industries and expert manhours that are spent on traveling and finally enhancing the productivity for the industries. Tests cases have verified that with AR, the resolution time could be significantly reduced, minimizing production stoppages by more than 50% of the time, which ultimately has a positive effect on a country's GDP. How much energy can be saved is predicted by the fact that 50% of all the world's business flights are replaced by one of the reality concepts and are estimated to amount to at least 50 Mton CO₂ per year. This figure is probably slightly higher as business trips also take place by other means of transport such as trains, buses, and cars. With today's volatile employees changing jobs more frequently, industry experts are becoming fewer and fewer. Since new employee stays for a maximum of 3–5 years per workplace, they will not stay long enough to become experts. Introducing an AI expert trained by today's experts, there is a chance that this knowledge can be maintained. [ABSTRACT FROM AUTHOR]

• The paper forecasts the Greek and the Hungarian day-ahead electricity price. • Forecasting is performed based on the ENTSO-E Transparency Platform data. • Rolling-window analysis is based on more than 1000 out-of-sample forecasting days. • Linearity bias is reduced with selected data mining and machine learning algorithms. • Impact of learning sample size on forecasting performance is examined. This paper evaluates the performance of alternative algorithms for day-ahead electricity price forecasting. Forecasting performance is assessed based on evidence from the Greek and Hungarian Power Market simulation. The electricity price formation process is simulated on a long time series spanning from January 2015 to September 2018. The EPF models are structured upon the explanatory variables that are available to the market participants before the exchange gate closure, through the publicly available ENTSO-E transparency platform. Relationships between the electricity spot price and explanatory variables are estimated by the selected econometric, data mining, and machine learning algorithms. The econometric autoregressive model with exogeneous explanatory variables is a benchmark model, as the other alternative approaches are used to overcome the linearity bias in the ordinary least squares estimator. We analyse the impact of a different training sample size as well as the impact of training on an hourly clustered sample on the forecasting performance. The support vector machine algorithm turned out to be the best alternative approach, with the lowest mean absolute error and statistically confirmed better forecasts compared to the benchmark econometric autoregressive model. The majority of the tested algorithms perform better with smaller training samples, whereas neural network based approaches prefer large training samples. Models with hourly clustered training samples have higher accuracy based on the Hungarian evidence, while hourly non-clustered training is a superior training method based on the Greek findings. [ABSTRACT FROM AUTHOR]