Search

Due to the low total cost of production, photovoltaic energy is a key component of installed renewable energy worldwide. However, photovoltaic energy is volatile in nature as it depends on weather conditions. This makes the integration, control, and operation of this type of energy difficult for network operators. Therefore, renewable energy forecasting is emerging as a key solution to effectively manage renewable energy in the power system. In this paper, we present a time series based energy forecasting system using a solar home database. We worked on Australian data recorded from 2010 to 2013. Three large databases have been combined to obtain better results. The objective is to test and compare the three prediction algorithms, namely: LSTM, FFNN and GRU. We also compared the performances via RMSE, MAE and R2_score. The results showed that neural networks accurately and reliably predict the energy consumed with R2_score up to 0.92 and MSE not exceeding 0.23, with a slight difference between the three models tested in this work.

This paper presents a groundbreaking control strategy for a bidirectional battery charger that allows power to be injected into the smart grid while simultaneously compensating for the grid’s reactive power using an electric vehicle battery. An artificial neural network (ANN) controller is utilized for precise design to ensure optimal performance with minimal error. The ANN technique is applied to generate sinusoidal pulse width modulation (SPWM) for a bidirectional AC–DC inverter, with the entire algorithm simulated in MATLAB Simulink.The core innovation of this study is the creation of the ANN algorithm, which supports grid compensation using electric vehicle batteries, an approach termed “vehicle-for-grid”. Additionally, the paper details the PCB circuit design of the system controlled by the DSP F28379D board, which was tested on a three-phase motor. The total harmonic distortion (THD) of the proposed ANN algorithm is approximately 1.85%, compared to the MPC algorithm’s THD of about 2.85%. This indicates that the proposed algorithm is more effective in terms of the quality of the power injected into the grid. Furthermore, it demonstrates effective grid compensation, with the reactive power effectively neutralized to 0KVAR in the vehicle-for-grid mode.

This paper investigates optimizing the power exchange between electric vehicles (EVs) and the grid, with a specific focus on the DC-DC converters utilized in vehicle-to-grid (V2G) systems. It specifically explores using model predictive control (MPC) in synchronous boost converters to enhance efficiency and performance. Through experiments and simulations, this paper shows that replacing diodes with SIC MOSFETs in boost converters significantly improves efficiency, particularly in synchronous mode, by minimizing the deadtime of SIC MOSFETs during switching. Additionally, this study evaluates MPC’s effectiveness in controlling boost converters, highlighting its advantages over traditional control methods. Real-world validations further validate the robustness and applicability of MPC in V2G systems. This study utilizes TMS320F28379D, one of Texas Instruments’ leading digital signal processors, enabling the implementation of MPC with a high PWM frequency of up to 200 MHz. This processor features dual 32-bit CPUs and a 16-bit ADC, allowing for high-resolution readings from sensors. Leveraging digital signal processing technologies and advanced electronic circuits, this study advances the development of high-performance boost converters, achieving power outputs of up to 48 watts and output voltages of 24 volts. Electronic circuits (PCB boards) have been devised, implemented, and evaluated to showcase their significance in advancing efficient V2G integration.

Although the EVRP (Electric Vehicle routing problem) has promising results on the environmental scale, its implementation has proved challenging. The difficulty of the EVRP resides in the limited driving range of the electric vehicles, combined with the significant charging time. While the charging cost is less than the cost of fuel, this charge time adds to the overall travel time and may overlap with customers’ time windows. All these factors increased the computational time exponentially and resulted in the need to overlook some constraints such as traffic congestion, road conditions, weather impact on energy consumption, and driving style, to name a few, in order to speed up execution time. While this method proved effective in accelerating the process of the EVRP, it did, however, render the approach unrealistic, as it steered far from real-life settings and made the approach unpredictable when facing dynamic and changing parameters. In this paper, we try to remedy this issue by proposing an approach in which we try to replicate real-life parameters such as heterogenous fleets, energy consumption, and infrastructure data. The objective of our approach was to minimize the total travel time, travel distance, energy consumed, and the number of vehicles deployed. To solve this problem, we propose a three-stages approach, in which the first stage consists of a newly developed dispatching approach where customers are assigned to vehicles. The second stage uses the genetic algorithm to find a set of optimal paths, and, finally, in the third stage, charging stations are inserted in the selected paths. Upon testing our approach on Solomon’s instances, our approach proved effective in finding optimal solutions in a reasonable time for five- to fifteen-customer datasets. However, when trying to solve larger datasets, the approach was slowed down by the extreme number of constraints it had to satisfy.

The intensive care units are a key element of patient flow, but due to high demand and an alternating rate of arriving patients, these units are often challenged by insufficient capacity, very high expenses, and in some cases, an unfair distribution of resources. Proper allocation of resources to match demand is, therefore, a vital task for many wards in these units. The patient bed assignment problem consists of managing in the best possible way a set of beds with equipment to be assigned to a particular type of patient. However, in real-world scenarios, constraints like a possible treatment trajectory are violated in most cases. In this paper, we present a new approach for solving patient bed assignment problems constrained by targets on survival function estimation, cost estimation, and possible treatment trajectory estimation for patients with cardiovascular diseases. For survival function estimations, we used the nave estimator and Kaplan-Meier, and for treatment effect estimations, we used logistic regression and T-learning. Estimations of the three components are used as weights in a genetic algorithm. This technique allows for the consideration of various constraints, which, unlike other techniques, allows for the selection of dominant solutions as solutions that satisfy dominant constraints. In addition, we demonstrate the robustness of our approach by testing the algorithms with multiple classes of patients, testing multiple sets of parameters, and comparing our results with several similar research studies showing the added value of working on this management axis in hospitals using the new approach to bed allocation.

Today’s environmental, energy and economic challenges require changes in the control strategies of HVACs’ systems, since they account for more than 60% of the building energy consumption. An optimal control law should be applied to reduce this consumption. To achieve this goal, a detailed description of the building is required, its construction components as well as the description of the occupants’ activities. A good knowledge of all building components is essential, especially the thermal characteristics of the building envelope, to build an energy model of the building. The dynamic thermal simulation tool, as EnergyPlus in this case, allows us to set up a huge database of the thermal behavior of our architectural model. This database was used for the training of an Artificial Neural Network model for modeling the thermal behavior of the building, with the aim of controlling the thermal comfort of the occupants, which means maintaining the temperature of the room within a setpoint temperature range. The developed control method, introduced in this work, reduced the energy consumption for cooling and heating, respectively, from 11.834 kWh to 9.025 kWh (27.34%) and from 4.631 kWh to 2.824 kWh (39.02%), compared to the On/Off control, during the 1 day of simulation.

The residential area represents a sector that consumes more electricity especially with the rapid urban growth and the transition towards the automation and electrification of several daily activities of human beings, namely: urban mobility and the residents’ indoor comfort. As a result, meeting this increased demand requires a proportional rise in fuel use to generate energy. This type of conventional power production has significant impacts on the environment. The eco-friendly alternative is the use of renewable energies in its distributed form on buildings. However, this solution gives rise to some issues related to energy management especially with the penetration of a new domestic device, namely: the plug-in electric vehicle. For this purpose, we propose a management system for a future household equipped with controllable electric loads and an electric vehicle equipped with a PV–Wind–Battery hybrid renewable system connected to the national grid. The proposed management system is based on a linear programming model with non-linear constraints solved with MATLAB toolboxes. The simulation is based on a database of meteorological conditions resulting from TRNSYS and processed to achieve a frequency of one hour. The system decisions provide switch control states of the connection architecture as well as the variation according to the V2H (vehicle to home), H2V (home to vehicle) and involved G2V (grid to vehicle) scenarios when grid comes into play during H2V mode.

Traditional agriculture has been the pillar of development on the planet for centuries. But with exponential population growth and increasing demand, farmers will need water to irrigate the land to meet this demand. Because of the scarcity of this resource, farmers need a solution that changes the way they operate. With the advent of new technologies, the notion of Agriculture 4.0 has become a reality to keep up with and meet the demand. With the addition of artificial intelligence and IoT through the collection and processing of agricultural data, decisions have become more and more precise to facilitate decision-making. This paper proposes an intelligent and flexible irrigation approach with low consumption and cost that can be deployed in different contexts. This approach is based on machine learning algorithms for smart agriculture. For this, we used a set of sensors (soil humidity, temperature, and rain) in an environment that ensures better plant growth for months, from which we collected data based on an acquisition map using the Node-RED platform and MongoDB. We used many different models based on the collected data: KNN, Logistic Regression, Neural Networks, SVM, and Naïve Bayes. The results showed that K-Nearest Neighbors is better with a recognition rate of 98.3% and a root mean square error (RMSE) of 0.12, compared to other models (LR, NN, SVM, NB). and towards the end, we provided a web application that brings together the various data emitted by the sensors as well as the prediction of our models to allow better visualization and supervision of our environment.

LoRa technology is being integrated into industrial applications as part of Industry 4.0 owing to its longer range and low power consumption. However, noise, interference, and the fading effect all have a negative impact on LoRa performance in an industrial environment, necessitating solutions to ensure reliable communication. This paper evaluates and compares LoRa’s performance in terms of packet error rate (PER) with and without forward error correction (FEC) in an industrial environment. The impact of integrating an infinite impulse response (IIR) or finite impulse response (FIR) filter into the LoRa architecture is also evaluated. Simulations are carried out in MATLAB at 868 MHz with a bandwidth of 125 kHz and two spreading factors of 7 and 12. Many-to-one and one-to-many communication modes are considered, as are line of sight (LOS) and non-line of Sight (NLOS) conditions. Simulation results show that, compared to an environment with additive white Gaussian noise (AWGN), LoRa technology suffers a significant degradation of its PER performance in industrial environments. Nevertheless, the use of forward error correction (FEC) contributes positively to offsetting this decline. Depending on the configuration and architecture examined, the gain in signal-to-noise ratio (SNR) using a 4/8 coding ratio ranges from 7 dB to 11 dB. Integrating IIR or FIR filters also boosts performance, with additional SNR gains ranging from 2 dB to 6 dB, depending on the simulation parameters.

Frequency modulated continuous wave (FMCW) radar is increasingly used for various detection and classification applications in different fields, such as autonomous vehicles and mining fields. Our objective is to increase the classification accuracy of objects detected using millimeter-wave radar. We have developed an approach based on millimeter-wave radar. The proposed solution combines the use of an FMCW radar, a YOLOv7 model, and the Pix2Pix architecture. The latter architecture was used to reduce noise in the heatmaps. We create a dataset of 4125 heatmaps annotated with five different object classes. To evaluate the proposed approach, 14 different models were trained using the annotated heatmap dataset. In the initial experiment, we compared the models using metrics such as mean average precision (mAP), precision, and recall. The results showed that the proposed model of YOLOv7 (YOLOv7-PM) was the most efficient in terms of mAP_0.5, which reached 90.1%, and achieved a mAP_0.5:0.95 of 49.51%. In the second experiment, we compared the models with a cleaned dataset generated using the Pix2Pix architecture. As a result, we observed improved performances, with the Pix2Pix + YOLOv7-PM model achieving the best mAP_0.5, reaching 91.82%, and a mAP_0.5:0.95 of 52.59%.

Overcrowded hospitals in its different process’s levels is a very common issue all over the globe. Research through different works have studied this problem from different angles. Most of those focused on a very specific part of the internal flows. This study presents a mathematical model aiming to reflect patient flow and resources usage. This model presents a new perspective on how it is possible to organize a hospital flow for inpatients of different pathologies. It is based on Non-Homogenous Discrete Time Markovian chains. It takes into consideration the patient’s pathology, survival function and the current beds distribution and discharges. To model the different variations of a patient’s flow, our work uses an actual case study to validate that our mathematical model based on Markov Chains can represent a real scenario of patients flow in a hospital and can predict the resources occupancy. We used as entry to this model the arrival distribution of patients as represented in the NHS dataset. Served patients and overflowing patients relate proportionally to admission rate based on how much time every diagnosis and preliminary treatment will take, then using the previous beds availability ration will permit, with knowledge of all present patients in the system, calculate the beds occupancy in the actual time. Only based on inputs from the NHS database, bed occupancy can be calculated using our model. Then compare this calculus with published data of the bed’s occupancy in the NHS. in the period between March 2020 and February 2021. The results of the simulation model were then compared to the dataset using the chi-square goodness of fit test. Non-homogenous discrete Markovian model, survival function, patient’s pathology, time-dependent, pathology, primary allocations, secondary allocations.

The goal of industrial revolutions is not only to improve and respond directly to the needs of industry in a productive side fact, but also to improve the standard of living of society and make life easier for customers. Therefore, economic growth should always go hand in hand with each industrial revolution. In this context, the increasing demand for electricity, the non-renewable nature of conventional sources and the rising price of fuel, are considered the constraints of the 21st century and need to be definitively managed in order to meet the constantly increasing demand for electricity from the population in the best possible conditions. Information and Communication Technologies (ICTs) are the foundation on which tomorrow’s innovative solutions are created. However, smart embedded systems and the internet are two major players driving ICT technologies forward. These technologies have impacted several sectors, where we find a crossroads of several industrial sectors: medical, manufacturing, automation, energy and others. The application these technologies in a smart grid framework with the integration of renewable energy sources and used for an optimization and efficiency mission let Internet of Things (IoT) to take the sustainable and industrial character and named Green Industrial Internet of Things (GIIoT). This last, seems to be a promoting axis for the development of future applications. In this paper, we present an overview of the impact of the Internet of Things on the evolution of applications in relation to the different revolutions, an inventory of the different industrial revolutions, an overview of the GIIoT as well as some applications and perspectives.

The increase demand for electricity and the non-renewable nature of fossil energy makes the move towards renewable energies required. However, the common problem of renewable sources, which is the intermittence, is overcome by the hybridization of complementary sources. Thus, whenever the load demand is not fully covered by the primary source, the second one will absolutely support it.Furthermore, the production, the interaction with the grid and the storage system must be managed by the grid-connected hybrid renewable energy system, which is the main objective of this paper. Indeed, we propose a new system of a grid-connected PV-battery, which can manage its energy flows via an optimal management algorithm. The DC bus source connection topology in our proposed hybrid architecture tackles the synchronization issues between sources when the load is powered. We consider in this work that choosing a battery discharge and charge limiting power provides an extension of the battery life. On the other hand, we simulated the dynamic behavior of the architecture’s various components according to their mathematical modeling. Following this, an energy management algorithm was proposed, and simulated using MATLAB/SIMULINK to serve the load. The results have shown that the load was served in all cases, taking into account the electrical behavior of the inhabitants as well as the weather changes on a typical day. Indeed, the load was served either by instant solar production between sunrise and sunset, or the recovery from sunset to 10pm, which could be a stored or injected energy without exceeding the 1000W per hour. Keywords: Renewable energy, PV-battery, Hybrid renewable system, Energy management, Hybrid architecture

This paper targets a major challenge of how to effectively allocate medical resources in intensive care units (ICUs). We trained multiple regression models using the Medical Information Mart for Intensive Care III (MIMIC III) database recorded in the period between 2001 and 2012. The training and validation dataset included pneumonia, sepsis, congestive heart failure, hypotension, chest pain, coronary artery disease, fever, respiratory failure, acute coronary syndrome, shortness of breath, seizure and transient ischemic attack, and aortic stenosis patients’ recorded data. Then we tested the models on the unseen data of patients diagnosed with coronary artery disease, congestive heart failure or acute coronary syndrome. We included the admission characteristics, clinical prescriptions, physiological measurements, and discharge characteristics of those patients. We assessed the models’ performance using mean residuals and running times as metrics. We ran multiple experiments to study the data partition’s impact on the learning phase. The total running time of our best-evaluated model is 123,450.9 mS. The best model gives an average accuracy of 98%, highlighting the location of discharge, initial diagnosis, location of admission, drug therapy, length of stay and internal transfers as the most influencing patterns to decide a patient’s readiness for discharge.

The industrial internet of things (IIoT) known as industry 4.0, is the use of internet of things technologies, via the Wireless Sensor Network (WSN), to enhance manufacturing and industrial processes. It incorporates machine learning and big data technologies, to allow machine-to-machine communication that have existed for years in the industrial world. Therefore, it is necessary to propose a robust and functional communication architecture that is based on WSNs, inside factories, in order to show the great interest in the connectivity of things in the industrial environment. In such environment, propagation differs from other conventional indoor mediums, in its large dimensions, and the nature of objects and obstacles inside. Thus, the industrial medium is modeled as a fading channel affected by an impulsive and Gaussian noise. The objective of this paper is to improve robustness and performances of multi-user WSN architecture, based on Discrete Wavelet Transform, under an industrial environment using conventional channel coding and an optimal thresholding receiver.