ANN-based optimized robust nonlinear MPPT algorithm for bidirectional quadratic converter in a PV system

As the demand for energy continues to rise, traditional fossil fuel reserves are depleting rapidly. Solar energy has garnered significant attention in sustainable energy systems due to its widespread appeal and potential. The output of photovoltaic systems fluctuates due to changes in factors like load requirements and atmospheric conditions hence the efficiency of photovoltaic energy is often hindered. To address this challenge, the pursuit of achieving optimal electricity production through the technique of maximum power point tracking (MPPT) is used. Due to the inherently nonlinear nature of PV systems, non-linear controllers are better suited for MPPT. The effectiveness of a PV system is intricately linked to the efficiency of the DC converters employed. The quadratic buck-boost converter boasts a broader voltage conversion range in contrast to the conventional buck-boost converter. This expanded range enhances its ability to adjust to fluctuating sunlight conditions, resulting in more effective extraction of power from PV panels. Artificial intelligence have emerged more valuable MPPT approaches and have gained widespread prominence. A robust controller is essential to guarantee effective maximum power point tracking, because the system’s nonlinearities are adapted. The aim of this paper is to employ the MPPT method utilizing Artificial Neural Network (ANN) and the non-linear condition-based super twisting sliding mode (CST-SMC) control to supervise the PV voltage reference up to a desired threshold value. A crucial application involves employing a pre-trained artificial neural network offline to establish the reference for peak power voltage. The experimental outcomes conducted through a hardware-in-the-loop (HIL) setup authenticate the real-time efficacy of the suggested controller.