Power Flow Controller Design for Permanent Magnet Synchronous Generator Based Wind Energy Conversion System Using an Adaptive Neuro-Fuzzy Inference System

Abstract

Because of the global energy crisis and environmental pollution problems affecting the world, the need for renewable energy sources is rapidly expanding. Amongst these renewable sources, the penetration of wind power is rising, which presents significant technological challenges for the development of electrical grid networks. For wind energy conversion systems, gearless permanent magnet synchronous generators are becoming more and more popular. The advancement of power electronics in the field of power systems has resulted in the development of new and more efficient power control technology. This thesis uses an Adaptive Neuro-Fuzzy Inference System (ANFIS) to control power flow of a Wind Energy Conversion System (WECS)-based on a Permanent Magnet Synchronous Generator (PMSG). As a benchmark for the performance of the fuzzy system, this includes the simulation and modeling of a wind speed, a wind turbine, a PMSG, an AC-DC uncontrolled rectifier, a boost converter, a DC-AC IGBT based PWM inverter, a filter, a conventional PI controller, a fuzzy logic controller, and an adaptive neuro-fuzzy inference system controller. The suggested controller uses ANN learning techniques for the issue of tuning a fuzzy logic controller and fine-tuning membership functions. The proposed control algorithms were validated using the simulation results obtained by MATLAB/SIMULINK. The result demonstrates that the proposed controller is highly effective, it has the best dynamic performance, and it is robust to line transmission parameter uncertainties when it is compared with the conventional PI controller and fuzzy logic controller. It is feasible. From the simulations, it can be observed that the overshoot given by the ANFIS controller is reduced to 0.53% from 10.404%, while the regulation time is reduced to 0.00685msec from 0.0585msec, and the settling time is reduced to 0.04248msec from 18.485msec from the PI controllers.