Evolutionary based Pareto optimization algorithms for bi-objective PV array reconfiguration under partial shading conditions

Abstract

During the daily operation of photovoltaic array, it easily faces the partial shading conditions resulted from the cloud shadow, dropping dust, etc. It will directly cause a lifetime reduction and a generation efficiency decrement for the photovoltaic array. To weaken the negative influence of partial shading condition, one of the most favoured ways is the photovoltaic array reconfiguration. However, the conventional photovoltaic array reconfiguration only aims to maximize the power output, which did not consider the lifetime and control complexity of switching devices. To fill up this gap, this paper constructs a new bi-objective optimization of photovoltaic array reconfiguration, which attempts to simultaneously maximize the output power and minimize the switching number. Consequently, it can dramatically reduce the switching control complexity while improving the generation efficiency, while the operation life of the switching devices can be lengthened. In order to find a high-quality Pareto optimal reconfiguration schemes, six frequently-used evolutionary multi-objective optimization algorithms are employed to solve this bi-objective optimization. The effectiveness of bi-objective optimization of photovoltaic array reconfiguration is tested on three scales of total-cross-tied photovoltaic arrays under four partial shading patterns. The simulation results show that the maximum power increment by the proposed technique is up to 26.6% against to that without optimization, while the average switch number decrement is up to 31.1% compared with the single-objective optimization algorithms.