Parameter Identification Algorithm for Grid-Connected Photovoltaic Power Generation System Based on Extended Kalman Filter

Abstract

Due to the rapid advancement of renewable energy sources, photovoltaic power generation systems (PVPGSs) have become increasingly prevalent in modern power systems. To mitigate potential degradation of system characteristics resulting from the proliferation of PVPGSs, power systems are imposing stricter regulation requirements on PVPGSs. However, the existing control methods for PVPGSs typically rely on prior knowledge of inverter control parameters, which are often not readily available or accessible. Inaccurate control parameters may constrain the efficacy of these control methods. Therefore, this paper presents an extended Kalman filter (EKF) based parameter identification algorithm for grid-connected PVPGSs to efficiently and accurately identify system parameters under disturbances. The mathematical model of the PVPGS is initially analyzed, followed by an analysis of the EKF algorithm and the proposal of an EKF-based PVPGS parameter identification method. Case studies indicate that the maximum errors of parameter identification results such as DC-side capacitance, proportion coefficient, and integration coefficient of PI regulator are less than 1.16% in different scenarios.