Non-dominated sorting genetic algorithm-II for robust multi-objective optimal reactive power dispatch

摘要

The concept of robust optimal solution is incorporated into multi-objective optimal reactive power dispatch (MORPD) for the consideration of uncertain load perturbations during system operations. Robust MORPD searches for solutions that are immune to parameter drifts and load changes. It uses information of load-increase directions to promote the stability of optimal solutions in the presence of load perturbations. Non-dominated sorting genetic algorithm-II (NSGA-II) is adopted to search for the robust Pareto solutions on a standard IEEE 118-bus system. The simulation validated the effectiveness of NSGA-II for robust MORPD. NSGAII obtained Pareto solutions over the trade-off surface. The experimental results also indicated that the robust Pareto solutions are comparatively less sensitive to load perturbations in their neighbourhoods and can maintain their objective values against uncertain load perturbations. Robust MORPD can provide optimal solutions with a higher degree of stability in the face of perturbations and can be more practical in reactive power optimisation of the real-time operation systems.