Maximum Power Point tracking implementation based on self-learning adaptive GA-Neural controller for standalone PV applications

• Innovative MPPT Strategies : The paper presents three novel Maximum Power Point Tracking (MPPT) strategies using Self Adapting Genetic Algorithm-based Neural Networks (SAGA-NN). • High Voltage Gain Control : The MPPT strategies are designed for controlling an uninsulated DC/DC boost converter with high voltage gain in a standalone photovoltaic (PV) system. • Hybrid Techniques : These strategies integrate Genetic Algorithms (GAs) and Multi-Layer Perceptron (MLP) neural networks, using a modified GA to optimize initial weights for improved accuracy. • Simulation and Experimental Validation : The effectiveness of the controllers is demonstrated through simulations in Matlab/Simulink and validated by real-world tests using a PV panel and emulator with a dSpace 1104 board. • Superior Controller Performance : The SAGA-ANN1 controller outperformed others, offering better overshoot, robustness, response time, and reduced oscillations under different weather and load conditions. This paper discusses a hardware implementation of three innovative Maximum Power Point Tracking (MPPT) strategies that utilize Self Adapting Genetic Algorithm based Neural Network (SAGA-NN) learning capabilities. These strategies are designed for controlling a high voltage gain, uninsulated DC/DC boost converter powered by a standalone photovoltaic (PV) system. Specifically, three novel control structures are introduced, employing different input variables and integrating three types of techniques to accomplish adaptive MPPT strategies. The Perturb and Observe (P&O) technique incorporates a combination of metaheuristic Genetic Algorithms (GAs) and a Multi-Layer Perceptron (MLP). A modified GA is used to determine an optimal set of initial weights, improving the MLP's accuracy. The effectiveness of the proposed controllers has been confirmed through simulations in the Matlab/Simulink environment and further validated by experimental tests using a solar PV module and a realistic solar PV emulator on a dSpace 1104 board. The experimental outcomes highlight the superior performance of the first proposed controller, SAGA-ANN1, in terms of overshoot, robustness, transient/steady-state performance, response time, and oscillations across varying weather and load conditions.