Design of Discrete-Value Passive Harmonic Filters Using Sequential Neural-Network Approximation and Orthogonal Array

This paper presents a method of combined sequential neural-network approximation and orthogonal array in the planning of passive harmonic filters. The purpose is to find the optimal locations for harmonic filters among existent capacitor. An orthogonal array is conducted to obtain the initial solution set for optimization. Then a back-propagation sequential neural network is trained to simulate the feasible domain for the optimal point. The purposes are to minimize the total harmonic distortion of voltages while satisfying harmonic standards. The proposed method was test on an 36-bus radial distribution system. Results obtained show that the proposed method is suitable for siting and sizing passive harmonic filters