Study On Hybrid Algorithm Of Power System Reactive Power Optimization

The power syttem reactive power supply and demand can not be coincide with each other in terms of geographic locations, which leads the reactive power flow variation in the grid and the buses voltage fluctuation. Voltage fluctuation may influence the voltage quality and the life of the electric installation, what is worse, the voltage fluctuation will likely result system fault and large area power failure. Therefore, reactive power opitimization means a lot to realizing the reactive power just distribution, reducing the reactive power flow in the grid, improving the security of the power system, promoting the voltage quality and improving economic of the system operation. This paper aims at combine two reactive power optimization algorithms to do the reactive power optimization hybrid method research.First of all, this paper based on the interior point nonlinear programming algorithm, set up the reactive optimization model by adopted the perturbed KKT conditions of the primal-dual interior point algorithm. By using the Newton algorithm to solve the KKT conditions function to get the optimum solution. Also the deduce process of the hession matrix and grade matrix was introduce in the paper.Second, Base on the actual operating control situation,power system reactive power control variables were divided into two type of variables, which is continuous variables and discrete variable. Interior point algorithm can not to treat with the discrete variable, so it needs to corporate with another algorithm. This paper proposes to use the particle swarm algorithm to treat the discrete variable. The fundamental principles and the relevant parameter of the particle swarm algorithm was introduced in the paper. In order to avoid the particle swarm converge early, the improvment of the exsiting particle swarm was made in this paper. The inertia weight can self-adapt its value according to the aggregation level during the swarm optimization progress , that play a balanced role of premature optimization.This paper proposed the combination algorithm by associating the interior point nonlinear programming algorithm and the particle swarm algorithm to solve the reactive power optimization problem.The proposed algorithm extemsive numerical simulation on test systems of IEEE14,IEEE30 nodes,have shown that the effectiveness of the proposed method。...