Parameter design optimization for Sandia frequency shift islanding detection method

Abstract

Islanding detection (ID) is a mandatory function for PV grid-connected converters. The traditional Sandia frequency shift method for ID becomes ineffective under certain parallel RLC loads. To alleviate this problem, a new methodology to optimize the frequency shift parameter Ks is proposed, and a novel non-detection zone (NDZ) mapping method is presented to map the NDZ influenced by each electrical parameter in this paper. An instability criterion relied on phase equilibrium principle algorithm is presented under specific loads to analyze the relationship between parameters and the reduction of NDZ. The proposed algorithm can keep the frequency of inverter output voltage deviating until the protection circuit is triggered with lower harmonic current. Single-phase photovoltaic grid-connected simulation system and experimental platform are constructed according to IEEE Std. 2000-929. Both the simulation and experimental results prove the robustness and effectiveness of the proposed parameter optimization algorithm.