Specialized tabu search algorithm applied to the reconfiguration of radial distribution systems

Abstract

This work presents a specialized tabu search algorithm applied to the problem of electric power distribution systems primary feeders’ reconfiguration. The specialization is related to two fundamental aspects of the tabu search algorithm. The first proposal eliminates the concept of a list of prohibited attributes and the aspiration criterion, but also avoids the possibility of revisiting a candidate solution so that cycling is avoided by maintaining a tabu list with all previously visited solutions. The second proposal is the possibility of restarting the search from the incumbent solution while avoiding paths that can be formed by revisiting candidate solutions. A new strategy based on Prim's algorithm generates a high-quality initial solution for the problem. Tests are conducted using the 33-, 84-, 118-, 136-, and 415-node test systems. The results demonstrate the effectiveness of the proposal for solving the reconfiguration problem since the best-known solution for each system is achieved within highly efficient execution times.