Squared Error Autocorrelation-Based VSS-LMS Control Algorithm for Grid-Integrated Solar Photovoltaic System

Abstract

The electrical power generated from the solar photovoltaic system is not constant due to the changing atmospheric parameters which are not desirable for reliable power supply and grid management. In this paper, a variable step size least mean square using squared error autocorrelation (VSS-SEA)-based control algorithm for a two-stage grid-integrated photovoltaic system is proposed. The value of weights is small during steady-state for low misadjustment level and increased during the transient state for fast convergence incremental conductance control for MPPT. Results are obtained for a 25-kW solar photovoltaic system using VSS-SEA under load unbalancing and varying solar irradiance for different loads and compared with the least mean square algorithm. Performance of the system using VSS-SEA is faster, smoother and more stable as compared to LMS under dynamically changing conditions. Moreover, the total harmonic distortion is found to be 1.76%, which is under IEEE-519 limits.