Service restoration in distribution systems using an evolutionary algorithm

This study proposes an efficient and computationally feasible solution approach based on evolutionary algorithms, to the distribution system restoration problem. In recent years, the enlargement and development of electrical power systems have introduced various problems related to the continuity of load feeding. When a fault occurs in a distribution system, the faulted area should be isolated from the system and as much load as possible has to be restored to this area. This is done by transferring de-energized loads from out-of-service areas through the distribution system without violating or disturbing the existing operation constraints. The solution of this problem, which can be defined as obtaining an optimum system configuration providing minimization of deenergized areas on the system after a fault occurred, is known as service restoration in distribution networks. The main concern of this problem is to minimize the impact of the outage on the system. This is achieved by finding an optimal operation configuration via changing the status of sectionalizing switches, taking into account the objectives and the constraints of the system. This makes the distribution system restoration problem a complex combinatorial optimization problem with multiple constraints. Therefore, to solve this problem and obtain feasible solutions of good quality, with an acceptable amount of computational effort, an evolutionary algorithm approach is proposed. The proposed evolutionary algorithm is applied to a radially configured distribution system model and analysis results are presented.