Methodology of solution for the distribution network reconfiguration problem based on improved harmony search algorithm

Abstract

The topology of power distribution systems can be modified by altering the status of sectional and tie switches for load management and protection, while maintaining the required network radiality. Several effective optimisation algorithms and graph theory-based techniques for system radiality detection have been proposed in specialised literature to solve the distribution network reconfiguration problem, usually seeking to achieve power loss reductions and voltage profile improvements. This work aims to contribute to this line of research and its novelties are composed of proposing as solution method a methodology based in an enhanced version of harmony search metaheuristic called improved harmony search algorithm, and a new simple and effective system impedance matrix based process to detect islanding of nodes as a strategy to meet radiality constraint. The proposed methodologies were applied to 33-bus, 84-bus and 118-bus distribution networks known from technical literature, aiming to total active power loss minimisation. As comparison between methods, a key focus in analysis of results is related to discuss some advantages and effectiveness of improved harmony search over original harmony search algorithm to attain optimal radial topologies.

A Novel Solution Method for the Distribution Network Reconfiguration Problem Based on a Search Mechanism Enhancement of the Improved Harmony Search Algorithm

“…In [25], the IHS algorithm is applied to the DSR problem, where a higher convergence speed compared to the basic HS is verified by the authors of the paper. This work aims to propose an improvement of the IHS algorithm with the introduction of the search mechanism known as Path Relinking.…”

“…For this reason, it is necessary to complement the verification of constraint (6) with the use of methods for detecting islanding, which are usually based on graph theory. In this work, we used the method proposed in [25] to test the radiality of each new configuration obtained in the process of finding the optimal solution of the DSR problem.…”

“…As discussed in [25], applying HS to the DSR problem may require too many iterations when the search space is too large (typically, large networks), such that the HS algorithm may get stuck momentarily in local minima or even fail to find the global minimum. To overcome these difficulties, improvements in the solution finding process of the HS algorithm were proposed in the literature.…”

“…Following a similar approach, the authors in [24] propose a dynamic DSR approach also considering battery energy storage systems and renewable resources; nonetheless, the application is carried out from the standpoint of the distribution system operator and is also limited to small and medium-size distribution systems. In [25], an Improved Harmony Search (IHS) algorithm is developed to solve the DSR problem. The authors also present an effective system impedance matrix based process to detect isolated nodes as a strategy to meet the radiality constraint.…”

“…This paper presents an efficient way of solving the DSR problem by means of an IHS algorithm. The main contribution of this paper is the improvement of the algorithm presented in [25]. In the proposed approach, a Path Relinking phase is added which accelerates the convergence of the algorithm.…”