Genetic algorithm based control for unbalanced low voltage networks

There always exist unbalance in low voltage networks. It has adverse effects like current increase and voltage drops below regulatory limits in the most loaded phase, presence of current in the neutral conductor, over-voltage problems in the least loaded phase, and feeder tripping due to neutral overcurrent. Classic solutions involve grid reinforcements. In this paper a control methodology that may be applied to any unbalanced network is proposed. A real unbalanced low voltage network was modeled and simulated with OpenDSS and Matlab. The proposed control based on genetic algorithm found the least number of phase swaps required to restore voltage violations. It resulted in percentage voltage unbalance factor improvement and 18% losses reduction.