Computer-based strategy for the restoration problem in electric power distribution systems

Abstract
Electric power distribution network restoration is an emergency activity in system control, a real-time environment. It takes place when the basic function of the electric utilities, to supply customers, fails due to faults in the network. Restoration of supply to the affected customers can be a large-scale combinatorial problem with the control operators under pressure to find a restoration solution and implement it as quickly as possible. Therefore, they need an effective computer-based restoration solution to aid them. The paper describes a modular restoration strategy designed to deal with realistic, large-scale outages caused by faults of any type in HV and MV distribution networks. It is based on heuristic search guided by expert knowledge. In searching for a restoration solution the strategy utilises the available spare loading capacity of both the HV and the MV networks as necessary. The algorithm observes the loading constraint in discrete time resulting in more affected load demand supplied in the case of partial restoration and a smaller number of switching operations in the general case. By giving preference to remotely operated switching devices over manually operated switching devices it brings about fast implementation of a restoration scenario. In the case of partial restoration it restores important customers with priority. To do that more efficiently, it considers a hierarchy of the important customers. The proposed restoration strategy can be an effective aid to control operators, however it is also an effective tool in training and in establishing network reconfigurations for maintenance work and in assisting system security analyses.