Fuzzy inference application to planning of distribution network switching

Optimal switching pattern planning is the main subject of distribution network reconfiguration to recover from a blackout or to avoid a blackout due to maintenance and construction. This is a combinatorial optimization problem, and an enumerative check of every pattern is necessary to obtain the optimal solution. In practice, successive heuristic search methods are adopted to obtain the quasi-optimal solution within a limited computation time. However, a very long computing time is still required when the problem size becomes slightly larger. The step-by-step nature of the methods results in many iterations, although it ensures satisfaction of the constraints. Reduction of the iterations is considered to be necessary to enhance processing speed. In this paper, “supply adequacy” is introduced as a new criteria index to guess the solution, which is obtained by fuzzy inference for each possible pair of source and load zone. The reconfiguration pattern is obtained at once or in a few processing steps by assigning to each zone the source with the highest adequacy. Numerical examples show that the proposed method can obtain the solution faster than the ordinary method, more remarkably with the increase of problem size.