PSO based multi-objective optimization of unbalanced lv distribution network by PV inverter control

Abstract

The rapid expansion of consumer-driven installations of residential photovoltaic (PV) systems causes serious power quality issues, e.g. notable voltage variations and unbalance, which not only detrimentally affects security and stability of distribution network operation, but also limits the number and capacity of further PV connections. Based on both reactive power management and real power curtailment of inverters, this study proposes a comprehensive PV control strategy to improve the performance of unbalanced three-phase four-wire low voltage (LV) distribution networks with high PV penetrations. The optimal combination of PV set points is determined by solving a multi-objective optimal power flow (OPF) problem that can simultaneously improve voltage magnitude and balance profiles while minimizing network loss, inverter loss associated with reactive power generation and the cost of real power generation curtailment. To reflect preferences on control objectives, the multi-objective problem is reformulated into an aggregated single-objective problem using weighted sum method and then solved by the global particle swarm optimization (PSO) in MATLAB. Detailed simulations are performed and analysed for a typical scenario of high PV penetration on a real three-phase four-wire unbalanced distribution network in Perth Solar City trial, Australia.