Reference point based non-dominated sorting approach for multi-objective optimization of power flow

This paper presents an evolutionary algorithm, reference point based multi-objective group search optimizer using non-dominated sorting approach (r-NSGSO), for optimal power flow problem with multiple objectives. First, the six objectives of optimal power flow are reduced to two objectives which represent the secure and economic indices, respectively. The r-NSGSO integrates the non-dominated sorting approach into the original multi-objective group search optimizer to get better converged and more evenly distributed solutions, and introduces a reference point provided by the decision maker to guide the search toward the Pareto-optimal region of interest. Simulation studies are undertaken on a ZDT test problem and the standard IEEE 30-bus system, and the results demonstrate the capability of r-NSGSO for the multi-objective optimization problems.