Pareto multiobjective optimization technique for reactive power planning

Reactive power planning is a constrained combinatorial multiobjective optimization problem. The problem is often converted into a constrained single-objective function to achieve a unique optimal solution. This article introduces a new method based on Pareto dominance incorporated into particle swarm optimization to deal with the reactive power planning problem as a multiobjective problem. The two conflicting and noncommensurate objective functions related to the cost and the active power losses are resolved. This method uses an external archive to preserve diversity and produce a well-distributed Pareto front. This approach considers the requirement of an acceptable voltage profile at each node; all load busses are candidates for capacitor installation and the capacitors are added in binary steps. The proposed approach evaluates all possible combinations of capacitor placement considering various cost function for capacitor banks. The results provide the designer with a wide set of optimal solutions (Pareto set) to choose from.