Design of eigenvalues sensitivity-based passive filter using genetic algorithm

Filters are used to overcome the problems of harmonics in distribution networks. Since the magnitudes of these harmonics depend to the system's eigenvalues, so, the filters are usually designed based on eigenvalues sensitivity. This method requires calculating the space state equations of the system. In this paper, a new method is proposed for filter design. In the proposed method, there is no need to calculate the space state equations of the system, but it just requires to optimizing the used genetic algorithm. The proposed filter is capable of eliminating the target harmonics during application. In addition, it is able to eliminate spontaneously some harmonics in other buses of the system. Also, applying the intelligent search solution for proposed filter not only facilitates the finding answer process but also increases the filter's flexibility. Moreover, it provides the power networks with easy change of filter structure with rapid and optimum responsiveness to probable changes of the network. The ability of the proposed filter is confirmed by the simulation results.