Planning passive filters using NSGA II for industry applications

Passive shunt harmonic filters are a convenient means of power quality preservation as well as effective reactive power compensators. The problem formulation for the optimization of passive filters has been addressed using various approaches as a single-objective or a multi-objective optimization problem. The present paper considers a new multi-objective formulation for the multiple passive filters selection and placement problem that includes the maximization of the reactive power compensation benefits accomplished by the passive filters as reactive compensators. While previous contributions solve the multi-objective problem by the minimization of a single objective function composite by several sub-objectives, the present paper use the Non-Dominated Sorting Genetic Algorithm (NSGA - II) for solving the problem. The algorithm is capable to determine the location and configuration of each filter (type, branch number, frequency tuning and quality factor). The presented algorithm find the best solution for all the buses simultaneously using many filters and also search the reactive power distribution by branch which offers the best results. The effectiveness of the proposed procedure is tested by solving a practical example.