Machine Learning-Based Adaline Neural PQ Strategy for a Photovoltaic Integrated Shunt Active Power Filter

This paper introduces novel techniques based on Machine Learning (ML) algorithms for a Photovoltaic integrated Shunt Active Power Filter performance improvement. The first goal is to design an efficient maximum power point tracking MPPT strategy in order harness the largest amount of energy possible. Thereby, new hybrid Support Vector Regression Perturb and Observe (SVM regression-P&O) algorithm proposed. SVM block improves speed by predicting initial duty cycle, whereas small fixed-step P&O ensures high accuracy. second purpose upgrade harmonics detection exploiting characteristics intelligent learning Adaline combined with ML algorithm. Therefore, regression-Adaline PQ designed. generates predicted weights Adaline, thus ensuring fast identification DC active component. In addition, ability this work rate parameter allows accurate extraction contrast Adaptive technique where performances are highly dependent chosen parameter. A comparative analysis various models carried out get best output prediction each regression block. Simulations have been performed confirm supremacy strategies over classical techniques. Finding exhibits significant decrease PV losses (up 99%), minor overshoot impressively extraction’s response time 98.8%), PVSAPF quality enhancement under online intermittent weather conditions variable nonlinear load.