Browsing by Author "Kheldoun, Aissa (supervisor)"

Now showing 1 - 6 of 6

Bearing faults classification of induction motor using advanced deep learning techniques
(Université M'hamed Bougara Boumerdès: Institue de génie electronic et electric, 2024) Djelouli, Seyyid Ahmed; Kheldoun, Aissa (supervisor)
This study investigates the application of advanced neural network models for bearing fault detection using vibration and current signals. Bearing faults in induction motors pose significant challenges to industrial operations, often leading to unexpected downtimes and increased main-tenance costs. The study explores the performance of Artificia lNeura lNetwork s(ANN) ,1D- Convolutional Neural Networks (1D-CNN), 1D-CNN with multi-kernel sizes, and Long Short-Term Memory (LSTM) models. Findings indicate that 1D-CNN and its multi-kernel size variant outper-form other models, achieving accuracies up to 99.95% under various load conditions for vibration data. The 1D-CNN multi-kernel size model’s ability to capture diverse features through different kernel sizes proved advantageous, reflectin g asignifica ntimproveme ntov erprevio usmethodologies that relied on extensive preprocessing.For the current signal dataset,Our recent finding ssurpas sall prior results, particularly in variable speed operation, where our work marks a pioneering effort. In our current signal dataset, the pinnacle of accuracy, reaching 99.88%, was attained through the application of the 1D-CNN model with the variable load operation dataset. This remarkable success highlights the effectivenes so fmergin g1D-CN Nwit hVariationa lMod eDecompositio n(VMD) ,en- abling the proficien tdecompositio no fsignal san dresolutio no fboundar yeffec ts tohand leintricate fault patterns. Despite encountering greater complexities in variable speed operation, our models persevered and achieved commendable accuracies. Notably, the 1D-CNN model achieved an accuracy of up to 99.36%. These results highlights the significan tachievemen tmad ei nterm so fdiagnosi si ninduction motors.
Control of stand-alone PV system under non-uniform irradiance
(2019) Zerrouki, Nihal; Dahman, Ghenima; Kheldoun, Aissa (supervisor)
Renewable energy sources play vital role in power generation, research in this area has grown rapidly in the last few years and the society is now aware of the harmful effects of fossil fuel on the environment and with the increased cost of fuel production. It is very important to look for alternative efficient, clean and cheap energy sources. Solar energy is usually considered as one of the most promising renewable energy sources. The power generation from the photovoltaic panels is subjected to varying environmental conditions such as temperature and irradiance which lead to a varying conversion efficiency. This necessitates an optimum use of the incident solar radiation. Since a PV unit supplies maximum electrical power only at a certain operating point, Maximum Power Point Trackers (MPPT) were developed to seek this optimal operation under changing light and load conditions. These methods would ensure an efficient and more reliable energy source. This project is intended to simulate and implement a GMPPT for stand-alone PV system in order to provide a fast, efficient tracking solution. PV generators develop a very complex power versus voltage characteristics. That is, under non-uniform isolation, the PV generators exhibit a curve with many power peaks. Identifying the appropriate peak during the operation is the goal of the MPPT for Proper and efficient operation of the PV system, therefore three algorithms are to be studied during this project which are the incremental conductance algorithm one of the most commonly used technique, the golden section search GSS algorithm and the particle swarm optimisation PSO algorithm, a comparison would be done by implementing the algorithms using MATLAB/SIMULINK to find the best MPPT algorithm that converges rapidly avoiding errors and obtaining the best efficiency.
Detection and classification of faults in photovoltaic modules
(2024) Mekhalif, Ibtihel; Kheldoun, Aissa (supervisor)
Solar photovoltaic systems are being widely used in green energy harvesting recently. At the same rate of growth, the modules that come to the end of life are growing fast. Therefore, rapid fault detection and classification of PV modules can help to increase the reliability of the PV systems and reduce operating costs. Inspection and maintenance of solar modules are important to increase the lifetime, reduce energy loss, and environmental protection. A combination of infrared thermography and machine learning methods has been proven effective in the automatic detection of faults in large-scale PV plants. However, so far, few studies have assessed the challenges and efficiency of these methods applied to the classification of different defect classes in PV modules. In this dissertation, an efficient PV fault detection and classification method is proposed to classify different types of PV module anomalies using thermographic images utilizing convolutional neural networks (CNN) and artificial neural networks (ANN). Eleven types of PV module faults such as cracking, diode, hot spot, offline module, and other faults are utilized. Several evaluation metrics were used toassess the performance namely accuracy, recall, precision, and F1 score. The testing accuracy was obtained as 91% for the detection of anomalies in PV modules and 91% to classify defects for four classes and 73% for twelve classes. In conclusion, the integration of advanced imaging techniques and machine learning algorithms presents a promising avenue for enhancing the reliability of PV systems. As the demand for clean and sustainable energy continues to grow, such innovations will be instrumental in meeting global energy needs while minimizing environmental impact. The advancements outlined in this thesis represent a significant step forward in the pursuit of more efficient and resilient renewable energy infrastructures.
Development of new maximum power tracking techniques for stand-alone PV system under nonuniform irradiance conditions
(Université M’Hamed BOUGARA de Boumerdes : Institut de génie electrique et electronique (IGEE), 2020) Belmadani, Hamza; Mellal, Sohaib; Kheldoun, Aissa (supervisor)
The overwhelming need to decarbonize the energy sector to peter out climate changes, and catch up with the increasing demand of energy, have paved the way to an immense deployment of renewables around the globe. Solar systems are used to convert sunlight that hits their panels into electrical energy via the photovoltaic effect. However, photovoltaics have a very low efficiency, and the generated power depends almost entirely on the amount of collected solar irradiance, temperature, the electrical load and the ambient circumstances that surrounds them. Since it is not possible to have a fixed stream of solar radiation or temperature, it is crucial to come up with effective means to tackle these problems. In this regard, Maximum power trackers are integrated with PV systems to cope with the dynamically fluctuating operating conditions, and keep the generated power as high as possible. This thesis focuses on maximum power point tracking (MPPT) in PV systems using soft computing techniques. Equilibrium Optimizer, Seagull Optimization and Slime Mould Algorithm are three novel metaheuristic techniques proposed in this project. Matlab and Simulink are used to simulate a standalone PV system driven by an MPPT controller and assess the three stated optimizers. The recommended techniques demonstrated outstanding results, under distinct insolation levels and complex shading conditions. To confirm their effectiveness, a comparative study on the basis of robustness, convergence time and efficiency, is carried out along with other well-known techniques: Particle Swarm Optimization (PSO), Whale Optimization (WOA), Grey wolf Optimization (GWO), Wind Driven Optimization (WDO) and the Grasshopper Optimization algorithm (GOA). Obtained results revealed that the proposed algorithms are either superlative or competitive in terms of both convergence speed and tracking efficiency.
Optimization of fuzzy controller parameters for the vector controlled induction motor drive
(Université M'hamed Bougara Boumerdès: Institue de génie electronic et electric, 2024) Aguenarous, Mohamed; Zidane, Houssam Eddine; Kheldoun, Aissa (supervisor)
Due to its competitive features, particularly ruggedness and cost-effectiveness ,the induction motor is the most widely used source of variable speed drives in the industry. Three common methods of control are employed based on the required performance level: scalar control, vector control, and direct torque control. The scalar controller is simple to implement but offer slimite dperformance ,whil evecto ran ddirec ttorqu econtro lar eused when high-performance speed control is required. However, both schemes typically rely on PID controllers, which are sensitive to the operating point and parameter variations. This project investigates the use of Fuzzy Logic Controllers (FLC) due to their ro-bust performance and model-free characteristics. Despite their advantages, the main drawback of FLCs is their complex tuning process, which limits their widespread utiliza-tion. To address this limitation, the project proposes optimizing the parameter selection process using meta-heuristic algorithms, specificall yParticl eSwar mOptimizatio n(PSO) and Grey Wolf Optimization (GWO). Three optimization schemes were employed: gains optimization, membership functions (MFs) optimization, and combined gains and MFsoptimization. Indirect Rotor Flux-Oriented Control (IRFOC) was used to carry out this study. The simulation results indicated that the system’s performance improved compared to classi- cal control techniques such as scalar control method. Furthermore, the proposed control scheme significantl yenhance dth esystem’ srobustness .Thi sprojec tacknowledge sthe challenges associated with the precise selection of fuzzy parameters and the optimization process. This work demonstrates that the implementation of Fuzzy Logic Controllers (FLCs) in industrial settings has led to significan tadvancement si ninductio nmoto rtech-nology. FLCs have enabled improved performance, increased robustness, and enhanced efficien cy int heoperati on ofinducti onmotor s,maki ngth e mamo reviab lea ndattractive option for various industrial applications.
PV Power forecasting using machine learning techniques
(2021) Cherchari, Abdelmalek; Bourouis, Ahmed; Kheldoun, Aissa (supervisor)
Due to the overwhelming challenge of catching up with the increasing demand of energy and the pressing need to greenify the energy sector to face the sensitive topics of climate changes and global warming, the importance of renewable energy sources experienced an impressive augment that is expected to continue. Hence Solar photovoltaic plants are widely integrated into most countries worldwide. either via grid-connection or stand-alone networks, as a result, forecasting the output power of solar systems, this constitutes the main challenge towards ensuring large-scale and seamless integration of photovoltaic systems to improve the accuracy of energy yield forecasts. However Photovoltaic (PV) power generation is prone to fluctuations and it is affected by different weather conditions. In this case, accurate forecasting provides the grid operators and power system designers with significant information to manage the power of demand and supply. This project aims to analyze and compare various machine learning based forecasting methods in terms of characteristics and performance. This comparative study of the models is done through error analysis. The accuracy is evaluated using historical weather data. In addition, this dissertation investigates the assessment of these models based on some well-known metrics. The obtained results show that some forecasting models for PV systems are more effective than others.