Exhaustive search and Multi-objective Evolutionary algorithm for single fault service restoration in a real large-scale distribution system

Abstract

A practical and efficient method for service restoration in distribution systems was developed and demonstrated on tests performed on the real and large-scale distribution system of Londrina city (Brazil). The method combines Multi-objective Evolutionary Algorithms with the tree encoding named Node-Depth Encoding and an Alarming Heuristic in order to find adequate service restoration plans for distribution systems with size from 3,860 buses and 632 switches to 30,880 buses and 5,166 switches, requiring running time less than 37 seconds for all the test cases. Moreover, the method requires no network simplification (as modeling a set of loads in a unique load point or using a relatively small set of switches instead of all switches) in order to generate adequate service restoration plans for those systems. This paper proposes a simple yet effective improvement by incorporating an exhaustive search as a first stage of that method, which guarantees the analysis of all possible initial service restoration plans (i.e. those requiring the minimal number of switching operations that can reconnect all the healthy out-of-service areas). Simulations results have shown the effectiveness of including the exhaustive search in that method.