Reactive Power Optimization And Planning Based On Particle Swarm Optimization Algorithm

Power system voltage and reactive power control is an effective measure to keeppower system safety and economic, and an important method to improve voltageequality. The purpose of reactive power optimization is to determine the optimalcontrol strategy and optimize sizes and locations of Var sources, so that the securityand economic operation of power systems can be maintained.From the view of mathematics, reactive power optimization and planningproblems are multi-constrained, non-convex, non-linear, which can be formulated as amixed-integer optimization problem (MINLP). This thesis presents a particle swarmoptimization (PSO) algorithm for the above problems. PSO algorithm is a multi-agentalgorithm essentially, and is easily employed to treat non-linear problems and find theoptimal solution.The computational flow of PSO technique has been described in this thesis, andthe selection and modification of the concerned parameters has been discussed. Theproposed approach utilizes the global and local exploration capabilities of PSO tosearch for the optimal settings of the control variables including the continuous anddiscrete variables. The process also shows its characters of parallel computing androbustness.In this thesis PSO has been examined and tested on the standard IEEE30, 57 bussystems, and also employed to solve reactive power planning of Anshan 220kV and66kV system. The results are fairly satisfactory,and also confirm the potential of PSOand show its effectiveness, robustness and superiority.