Constrained robust multiobjective optimization for reactive design in distribution systems

This paper presents a new formulation includingrobustness of solution of constrained multiobjective design orreactive power compensation. The algorithm used foroptimization is the NSGA-II (Non dominated Sorting GeneticAlgorithm II) with a special crowded comparison operator forconstraints handling. The need for including the issue ofrobustness of solutions derives from the simple observation thatloads are uncertain in distribution systems and their estimation isoften affected by errors. In design problems it is desirable toconsider the loads with a certain range of variation. In this paperthe NSGA-II algorithm is applied to efficiently solve the issue andthe solutions attained considering stochastic loads are comparedto those considering precise loads.