Dependable Parallel Multi-population Global-best Brain Storm Optimization with Differential Evolution strategies for Distribution System State Estimation using Just-in-time Modeling and Correntropy in Power Systems

摘要

This paper proposes dependable parallel multi-population global-best brain storm optimization with differential evolution strategies (DPMP-GBSODE) for distribution system state estimation (DSSE) using just-in-time (JIT) modeling and correntropy. In electric power distribution systems, DSSE is utilized by power utility operators to grasp whole distribution system conditions such as voltages and currents. By applying JIT modeling and correntropy to DSSE problems, voltages and currents can be correctly estimated even if false measurement values (outliers) are measured. Considering equipment of the distribution systems and penetration of renewable energies (REs), it is necessary that an evolutionary computation technique with parallel and distributed processing (PDP) is applied to the DSSE problems. When some computational processes are distributed by PDP in server systems of distribution automation systems, some calculation results from the distributed computational processes may not be returned because of various congestions of the processes. Therefore, appropriate estimation results should be obtained even if the congestions occur (dependability). The proposed DPMP-GBSODE is verified to improve dependability and computation time for the DSSE problem in comparison with conventional method even if faults of the processes occur and the outliers are measured.