Integrated database approach in multi-objective network reconfiguration for distribution system using discrete optimisation techniques

Reconfiguring the link between buses is a crucial task to enhance the distribution system performance. Reconfiguration is a complex combinatorial process due to numerous feasible solutions. Therefore, to consistently find global optimum solutions within a short span of time is a challenging task. One of the factors that cause time consumption in finding optimal network configurations is the elimination of non-radiality network solutions during the optimisation process. To address this issue, this work proposes to store pre-determined network radiality solutions in a database. These sets of solutions are used in the network reconfiguration optimisation by a discrete evolutionary programming and a discrete evolutionary particle swarm optimisation techniques. These optimisation methods are based on a multi-objective problem which minimises power loss, voltage deviation, and a number of switching actions. Moreover, the quality of the solutions is measured in terms of computational time and consistency. To demonstrate the efficiency of the proposed technique, a comparative assessment is carried out on 33-bus and 118-bus distribution systems. It is found that the proposed technique outperforms other existing methods in terms of quality of the solutions.