NeuroPnO+ MPPT for a String of 5 PV Panels
September 2024 - May 2025
In my final year, I undertook a project focused on developing a novel Artificial Neural Network (ANN)-based Maximum Power Point Tracking (MPPT) system for photovoltaic (PV) energy applications. My project explored how intelligent, data-driven models, specifically ANN architectures, can adapt more efficiently than conventional methods, ensuring PV systems consistently opearate at their optimal power point.
I used MATLAB/Simulink with Simscape Electrical to design and simulate PV systems. This allowed me to test power converters under varying irradiance levels and temperature, giving me practical insight into how PV systems behave in real conditions.
I explored both conventional and AI-based MPPT methods. Traditional algorithms like Perturb & Observe (P&O) and Particle Swarm Optimisation (PSO) were compared againt my ANN-based P&O controller, highlighting how intelligent methods adapt faster to changing environments.
I developed and trained a Feed Forward Artificial Neural Networks to predict the optimum Global Maximum Power Point (GMPP) voltage more accurately. This approach improved response speed and efficiency, while also sharpening my skills in AI, research, and data analysis. I've shared my full thesis along with the MATLAB/Simulink models and resources I built during this project. You can check them out below:
