MPPT algorithm based on Multiple Linear Regression model for Solar PV systems

Abstract

Currently, there are numerous maximum power point tracking (mppt) algorithms implemented in systems powered by photovoltaic (PV) panels. These systems are indirectly affected by sudden changes in weather conditions, making them unstable and inefficient. This work presents a model to establish the optimal working point of photovoltaic panels to be used in installations sensitive to changes in environmental conditions, such as solar pumping, to obtain maximum power in any temperature and irradiance condition of the panels. The model reduces the response time of the system because it also has a control system implemented whose setpoint value is the one provided by the algorithm itself. This is possible with the help of external temperature and solar radiation sensors that provide these values in real time. The setpoint is the working voltage target for the panels (Vmpp), for which the system does not become unstable despite the oscillations allowed by the control system. The algorithm is based on a model that consists of a Multiple Linear Regression through which the entire P-V curve of any photovoltaic panel can be represented for any temperature and radiation.The characteristics of the model allow a control system design that makes the system more efficient.