Optimal design of discrete-value passive harmonic filters using sequential neural-network approximation and orthogonal array

This paper presents a method for combining sequential neural-network approximation and orthogonal arrays (SNAOA) in the planning of large-scale passive harmonic filters. An orthogonal array is first conducted to obtain the initial solution set. The set is then treated as the initial training sample. Next, a back-propagation sequential neural network is trained to simulate the feasible domain for seeking the optimal filter design. The size of the training sample is greatly reduced due to the use of the orthogonal array. In addition, a restart strategy is also incorporated into the SNAOA so that the searching process may have a better opportunity to reach a near global optimum. To illustrate the performance of the SNAOA, a practical harmonic mitigation problem in a chemical plant is studied. The results show that the SNAOA performs better than the original scheme and satisfies the harmonic limitations with respect to the objective of minimizing the total demand distortion of harmonic currents and total harmonic distortion of voltages. Filter loss, reactive power compensation, and the constraints of individual harmonics are also considered. Additional results related to SNAOA are also reported and discussed as well.