Reactive power optimization using agent-based grid computing

As a novel computation mode, the methodology of grid computing can exploit geographically and organizationally distributed computational resources for solving reactive power optimization problems. A grid-computing architecture for reactive power optimization is designed herein. Considering the characteristics of reactive power, a large-scale power system can be divided into many small-scale subsystems according to administrative domain and network structure. The optimization subproblems for corresponding subsystems can be worked out autonomously by agents, which are wrapped into grid services. The agents are diverse in objective functions, optimization algorithms and power flow algorithms. Grid computing technology is adopted to integrate the heterogeneous computational resources, which are distributed to different subsystems and dominated by different control centers. The simulation results of a practical power system show the feasibility of the proposed method