Support Vector Machine-Based PI-controlled MPPT for grid-connected Photovoltaic Systems

Control and power system engineering requires precise control action for achieving maximum efficiency. Support vector machine has a proven machine learning algorithm in achieving less error between actual and required output. One such area where SVM is applied with Solar PV system is Maximum Power Point Tracking. In MPPT enabled solar PV system, SVM is used to control the output duty cycle of the converter in order to extract maximum power from Solar PV system. In the proposed method a SVM MPPT algorithm has been proposed alongwith P and O method to generate the reference training dataset for duty cycle prediction. The proposed controller will reduce the prediction time and thereby achieving the faster response time for convergence. The paper also emphasis on the reliable control mechanism for effective integration of solar PV with the grid using Matlab Simulink model.