Search

The need for an efficient power source for operating the modern industry has been rapidly increasing in the past years. Therefore, the latest renewable power sources are difficult to be predicted. The generated power is highly dependent on fluctuated factors (such as wind bearing, pressure, wind speed, and humidity of surrounding atmosphere). Thus, accurate forecasting methods are of paramount importance to be developed and employed in practice. In this paper, a case study of a wind harvesting farm is investigated in terms of wind speed collected data. For data like the wind speed that are hard to be predicted, a well built and tested forecasting algorithm must be provided. To accomplish this goal, four neural network-based algorithms: artificial neural network ( ANN ), convolutional neural network ( CNN ), long short-term memory ( LSTM ), and a hybrid model convolutional LSTM ( ConvLSTM ) that combines LSTM with CNN , and one support vector machine ( SVM ) model are investigated, evaluated, and compared using different statistical and time indicators to assure that the final model meets the goal that is built for. Results show that even though SVM delivered the most accurate predictions, ConvLSTM was chosen due to its less computational efforts as well as high prediction accuracy., Competing Interests: The authors declare that they have no conflicts of interest., (Copyright © 2020 Mariam Ibrahim et al.)

The intermittent and stochastic nature of Renewable Energy Sources (RESs) necessitates accurate power production prediction for effective scheduling and grid management. This paper presents a comprehensive review conducted with reference to a pioneering, comprehensive, and data-driven framework proposed for solar Photovoltaic (PV) power generation prediction. The systematic and integrating framework comprises three main phases carried out by seven main comprehensive modules for addressing numerous practical difficulties of the prediction task: phase I handles the aspects related to data acquisition (module 1) and manipulation (module 2) in preparation for the development of the prediction scheme; phase II tackles the aspects associated with the development of the prediction model (module 3) and the assessment of its accuracy (module 4), including the quantification of the uncertainty (module 5); and phase III evolves towards enhancing the prediction accuracy by incorporating aspects of context change detection (module 6) and incremental learning when new data become available (module 7). This framework adeptly addresses all facets of solar PV power production prediction, bridging existing gaps and offering a comprehensive solution to inherent challenges. By seamlessly integrating these elements, our approach stands as a robust and versatile tool for enhancing the precision of solar PV power prediction in real-world applications. [ABSTRACT FROM AUTHOR]

Solar photovoltaic (PV) systems, integral for sustainable energy, face challenges in forecasting due to the unpredictable nature of environmental factors influencing energy output. This study explores five distinct machine learning (ML) models which are built and compared to predict energy production based on four independent weather variables: wind speed, relative humidity, ambient temperature, and solar irradiation. The evaluated models include multiple linear regression (MLR), decision tree regression (DTR), random forest regression (RFR), support vector regression (SVR), and multi-layer perceptron (MLP). These models were hyperparameter tuned using chimp optimization algorithm (ChOA) for a performance appraisal. The models are subsequently validated on the data from a 264 kWp PV system, installed at the Applied Science University (ASU) in Amman, Jordan. Of all 5 models, MLP shows best root mean square error (RMSE), with the corresponding value of 0.503, followed by mean absolute error (MAE) of 0.397 and a coefficient of determination (R2) value of 0.99 in predicting energy from the observed environmental parameters. Finally, the process highlights the fact that fine-tuning of ML models for improved prediction accuracy in energy production domain still involves the use of advanced optimization techniques like ChOA, compared with other widely used optimization algorithms from the literature. [ABSTRACT FROM AUTHOR]

Previous theoretical research efforts which were validated by experimental findings demonstrated the thermo-economic benefits of the hybrid concentrated photovoltaic-thermoelectric (CPV-TE) system over the stand-alone CPV. However, the operating conditions and TE material properties for maximum CPV-TE performance may differ from those required in a standalone thermoelectric module (TEM). For instance, a high-performing TEM requires TE materials with high Seebeck coefficients and electrical conductivities, and at the same time, low thermal conductivities (k). Although it is difficult to attain these ideal conditions without complex material engineering, the low k implies a high thermal resistance and temperature difference across the TEM which raises the PV backplate's temperature in a hybrid CPV-TE operation. The increased PV temperature may reduce the overall system's thermodynamic performance. To understand this phenomenon, a study is needed to guide researchers in choosing the best TE material for an optimal operation of a CPV-TE system. However, no prior research effort has been made to this effect. One method of finding the optimum TE material property is to parametrically vary one or more transport parameters until an optimum point is determined. However, this method is time-consuming and inefficient since a global optimum may not be found, especially when large incremental step sizes are used. This study provides a better way to solve this problem by using a multiobjective optimization genetic algorithm (MOGA) which is fast and reliable and ensures that the global optimum is obtained. After the optimization has been conducted, the best performing conditions for maximum CPV-TE energy, exergy, and environmental (3E) performance are selected using the technique for order performance by similarity to ideal solution (TOPSIS) decision algorithm. Finally, the optimization workflow is deployed for 7000 test cases generated from 10 features using the optimal machine learning (ML) algorithm. The result of the optimization chosen by the TOPSIS decision-making method generated an output power, exergy efficiency, and CO2 saving of 44.6 W, 18.3%, and 0.17 g/day, respectively. Furthermore, among other ML algorithms, the Gaussian process regression was the most accurate in learning the CPV-TE performance dataset, although it required more computational effort than some algorithms like the linear regression model. [ABSTRACT FROM AUTHOR]

Unmanned aerial vehicles (UAVs) have been highly enhanced in the last decade, targeting the deployment of UAV technology in different new applications in many fields. However, the motion control of the UAV during autonomous flight is still under development, especially in unstructured or dynamic environments. This paper presents an intelligent motion control framework using artificial potential field neural networks (APF-NNs) for UAVs in dynamic 3D environments. This approach is developed using potential field models for repulsive and attractive forces. Using the control commands of the APF-NN intelligent system, the UAV changes its planner position and altitude simultaneously to perform the path planning task. In other methods, the UAV only changes its planner positions for path planning task. The presented work is validated using simulations and field experiments. The proposed approach is outperforming existing fuzzy logic and classical potential field-based methods with a 75% and 86% reduction in computational time, 22% and 35% increase in inside success rate, 1.5% and 14.7% increase in the speed and 6.2% and 11% reduction in total traveled distance, respectively. [ABSTRACT FROM AUTHOR]

An artificial neural network (ANN) extracts knowledge from a training dataset and uses this acquired knowledge to forecast outputs for any new set of inputs. When the input/output relations are complex and highly non-linear, the ANN needs a relatively large training dataset (hundreds of data points) to capture these relations adequately. This paper introduces a novel assisted-ANN modeling approach that enables the development of ANNs using small datasets, while maintaining high prediction accuracy. This approach uses parameters that are obtained using the known input/output relations (partial or full relations). These so called assistance parameters are included as ANN inputs in addition to the traditional direct independent inputs. The proposed assisted approach is applied for predicting the residual strength of panels with multiple site damage (MSD) cracks. Different assistance levels (four levels) and different training dataset sizes (from 75 down to 22 data points) are investigated, and the results are compared to the traditional approach. The results show that the assisted approach helps in achieving high predictions' accuracy (<3% average error). The relative accuracy improvement is higher (up to 46%) for ANN learning algorithms that give lower prediction accuracy. Also, the relative accuracy improvement becomes more significant (up to 38%) for smaller dataset sizes. [ABSTRACT FROM AUTHOR]

This paper proposes computational models to investigate the effects of dust and ambient temperature on the performance of a photovoltaic system built at the Hashemite University, Jordan. The system is connected on-grid with an azimuth angle of 0° and a tilt angle of 26°. The models have been developed employing optimized architectures of artificial neural network (ANN) and extreme learning machine (ELM) models to estimate conversion efficiency based on experimental data. The methodology of building the models is demonstrated and validated for its accuracy using different metrics. The effect of each parameter was found to be in agreement with the well-known relationship between each parameter and the predicted efficiency. It is found that the optimized ELM model predicts conversion efficiency with the best accuracy, yielding an R2 of 91.4%. Moreover, a recommendation for cleaning frequency of every two weeks is proposed. Finally, different scenarios of electricity tariffs with their sensitivity analyses are illustrated. [ABSTRACT FROM AUTHOR]

[Display omitted] • Comprehensive performance analysis & forecasting of a WWTP in Jordan is performed. • The plant's performance is analyzed by resorting to various statistical methods. • A data-driven approach to predict plant production 1-month in advance is proposed. • SVM emerges as top performer, showcasing potential of hybrid forecasting approach. • Such analysis offers valuable insights to enhance the plant's operability. This study presents a comprehensive performance and forecasting analysis of the As-Samra wastewater treatment plant (WWTP) in Jordan, with two main objectives. Firstly, a thorough evaluation of the plant's performance is conducted. The analysis involves independently assessing historical operational conditions, plant production, and their statistical correlations using various statistical techniques. The second objective focuses on developing a data-driven forecasting approach to predict the plant's production one month in advance, using multiple machine learning models. The results highlight the effectiveness of principal component analysis (PCA) in simplifying operational data, revealing distinct operational clusters, and identifying seasonal production patterns while showing correlations between operational conditions and overall power production. The support vector machine (SVM) forecasting model emerged as the top performer, showcasing the potential of a hybrid forecasting approach. The findings offer valuable perspectives for enhancing operational efficiency, refining production planning, and ultimately improving the environmental impact of the plant. [ABSTRACT FROM AUTHOR]

[Display omitted] • Developed a universal method to estimate monthly tilted irradiance. • Tested three machine learning models and a deep learning model for 740 locations. • K-nearest neighbors model achieved the lowest RMSE and MAE. • Demonstrated enhanced prediction accuracy over the traditional isotropic model. • Data-driven models outperform in regions with consistent climate conditions. In this study, a universal method for estimating monthly solar radiation on tilted surfaces from horizontal measurements was developed and validated, targeting the optimal design and performance of solar energy systems. Using data from the Photovoltaic Geographical Information System (PVGIS) across 740 global locations, three machine learning models—linear regression, random forest, and k-nearest neighbors (KNN)—along with a multilayer perceptron deep learning model, were evaluated. The dataset spanned tilt angles from 0° to 90° at 1° increments and 11,511,773 rows with six input features for predicting irradiance on any tilt angle. The KNN model was selected for further testing analysis due to its superior accuracy, achieving the lowest root mean square error (RMSE) of 1.42 kWh/m2 and mean absolute error (MAE) of 0.7 kWh/m2 on the validation dataset. Testing was conducted with 16 years of data from five locations not included in the training dataset, obtained from PVGIS and Solcast. Results show that the KNN model consistently outperformed the traditional isotropic model in four out of five cities using PVGIS data and in two using Solcast data, especially excelling in regions with stable climates and consistent irradiance profiles. Moreover, better prediction performance was observed with PVGIS data compared to Solcast data across all five cities, with PVGIS data yielding an RMSE of 3.03 kWh/m2 and an MAE of 2.41 kWh/m2, while Solcast data exhibited a higher RMSE of 5.22 kWh/m2 and MAE of 4.14 kWh/m2. Finally, these results advocate for a shift towards data-driven methodologies highlighting their enhanced reliability. [ABSTRACT FROM AUTHOR]