Optimal planning of harmonic filters on distribution systems by chance constrained programming

Considering the random nature of harmonic sources in distribution systems, a chance constrained programming model for optimal planning of harmonic filters is presented in this paper. The locations and parameters of active and passive harmonic filters are determined simultaneously to minimize the total investment costs, while the harmonic voltage distortion limits and filter capacity constraints are satisfied with predetermined confidence levels. A stochastic simulation based genetic algorithm is used for solving this model. Test results on an 18-bus distribution system show that the proposed approach is more practical, efficient and economical than the conventional methods. (C) 2003 Elsevier B.V. All rights reserved.