NSGAII optimization for single phase passive filter allocation in distribution systems

This work presents a methodology for passive filter allocation within an electrical distribution system aiming to minimize the number of nodes that exceed harmonic distortion limits and investment costs. Harmonic distortions are usually mitigated by filters, but their allocation in the system requires attention. Depending on the filters location, harmonic distortion could increase at other nodes. The methodology of this study is based on harmonic flow simulation in the Alternative Transient Program (ATP/EMTP) as well as the NSGA II optimization algorithm simulated on Matrix Laboratory (MATLAB). The filters are located using single-phase units, in order to obtain solutions for unbalanced distribution systems with single-phase and two-phase circuits, while also attending features of single-phase inverter connected Distributed Generation (DG). Different scenarios are generated by MATLAB and simulated on ATP. The optimization problem is non-linear and multi-objective. One objective is to minimize the number of nodes exceeding distortion limits and the other is to minimize filter costs. The results show that the optimal installation of single-phase filters gives better results than the installation of three-phase filters at the critical nodes. The proposed methodology allows the user to choose the best option between the number of nodes exceeding limits and the investment cost.