Design and Implementation of Efficient MPPT Controllers based on SDM and DDM using Backstepping Control and SEPIC Converter

Abstract

This paper presents a design and hardware implementation of Maximum Power Point Tracking (MPPT) algorithms for photovoltaic panels. Two MPPT techniques have been presented which are SDM-BSC and ANNDDM-BSC. In the SDM-BSC control, the reference voltage that guarantees the maximum power point is generated by an analytical equation. The nonlinear backstepping control was introduced to track the reference voltage generated with the single diode modeling (SDM) by adjusting the duty cycle of a SEPIC converter. In order to improve the maximum power point tracking performance, the (SMD) is replaced by an artificial neural network (ANN) trained by a PV modeling based on DDM. The ANN is trained using double diode modeling (DDM), by varying the temperature and irradiance and computing the maximum voltage in every condition, thus a database of about ten photovoltaic module operating scenarios has been developed, which includes the temperature, irradiance and the maximum voltage delivered under every operating conditions. The ANN was designed with two inputs, which are temperature and insolation, and with a single output, which is the optimum voltage. The simulation results in the Matlab software showed excellent stability and speed compared to the InC-BSC and PO-BSC methods. And a power gain of about 0.3 W compared to the MSD-BSC method. The experimental setup has shown that the two proposed methods can be easily implemented and confirms the superiority of the ANNDD-BSC method.