A novel method for distribution system feeder reconfiguration using black-box optimization

This paper presents a novel method for distribution system feeder reconfiguration. The problem is formulated as a black-box optimization problem such that it can be solved by Mesh Adaptive Direct Search (MADS) algorithm. Given feeder reconfiguration candidates, a distribution system software simulator- OpenDSS is used to calculate detailed power flow solutions to provide various operating outputs such as system losses, line currents, and bus voltages. The radial topology constraints are maintained based on graph theory analysis, and two versions of constraints are formulated for binary programming and integer programming. The proposed method is a general framework, which is not restricted by the type of objectives, and OpenDSS enables detailed modeling of any distribution system. Test of the method on the IEEE 123 node feeder with distributed generators shows promising results, and the two versions of radial topology constraints are compared.