Optimization of photovoltaic pumping system using neuro fuzzy inference system ANFIS control technique

Abstract

In recent years, artificial intelligence has become increasingly used due to the development of microcontrollers. In this paper, we propose an intelligent technique that employs the adaptive neuro-fuzzy inference system (ANFIS). We use this approach to improve the conventional direct torque control (DTC), which relies on a PI controller for the induction machine, and to enhance the conventional MPPT control based on the Perturb and Observe algorithm. The overall goal is to improve the performance of the photovoltaic pumping system. In this work, we apply ANFIS control to maximum power point tracking (MPPT-ANFIS). Additionally, we simultaneously optimize the efficiency of the DTC by applying ANFIS control (DTC-ANFIS). We present the results by comparing the photovoltaic pumping system using ANFIS control with the conventional photovoltaic pumping system, using MATLAB/Simulink. The results show that ANFIS control significantly improves the photovoltaic system compared to the conventional control, offering excellent dynamic performance of the induction motor and better utilization of photovoltaic solar energy. However, the ANFIS has some drawbacks, such as high computational time consumption and challenges in implementing a database.