Distribution system service restoration by dynamic programming considering switch characteristics

Abstract

The objective of the proposed research is to develop a systematic approach of finding the optimal switching sequence leading to optimal post-fault topology in distribution system restoration (DSR) problem. The purpose of DSR is to reconfigure the topology of distribution system through switching actions to restore power to customers subject to system faults. Many existing methods formulate DSR problems as single-step optimization problem where the only control devices are remotely controllable sectionalizers. The problem with this approach is that when various kinds of switches are present in the distribution system, certain sequence of switching actions need to be followed based on the different characteristics of the switches, which is not considered by the single-step optimization formulations. To the best of our knowledge, there hasn't been research reported in the literature that considers the characteristics of switches and couples the constraints in the optimization problem to derive feasible switching sequences in a systematic and mathematically rigorous way. By formulating the DSR problem as a dynamic programming (DP) problem, the solution can be found in a systematic way with guaranteed optimality. Switch current limits as well as system operational constraints are considered in the formulation. Scalability to larger systems and related computational issues are discussed.