A fast and effective EA for service restoration of compensated distribution networks

An effective service restoration (SR) strategy for distribution network plays a key role in improving service reliability and enhancing customers´ satisfaction. Several evolutionary algorithms (EAs) have been developed in order to solve this problem. However, most of them do not include compensators in the plan. Others use some simplifying assumptions in order to limit the search space, which degrade the quality of obtained results. In addition, EAs suffer from high computational burden especially when improper modeling and operators are used. This paper proposes an integrated framework for EA in order to solve the SR problem which employs an efficient modeling of distribution network and simultaneously optimizes the network´s configuration, the amount of load shedding and compensators´ settings using special operators. Furthermore, a new technique has been proposed which limits the search space to only the area affected by the fault(s). Simulations prove efficiency of the proposed method by providing fast and high quality results.