Reactive Power Compensation Of Distribution Network Based On Particle Swarm Optimization

Optimal planning of reactive power compensation is of great significance in planning design of electric power system. In this paper, the optimal planning of reactive power compensation with tree shape electric power system is discussed.The mathematical model of optimal planning reactive power compensation for 10kv distribution power system is been established in this paper. In the objective function, the sum of active power lose cost and new adding VAR equipment investment which has been divided in each year are considered, and reactive power balance, satisfaction voltage quality are the constraints conditions. So, the optimal planning result can ensure good qualities and economical benefits of electric energy. In the model considers three kinds of different load operation conditions-light, normal and heavy are considered. This makes the model more comply with the actual situation and avoid the one-sidedness of single load operation condition.In order to make the results more precise and the algorithm more simple, the model is decomposed into three sub-models under light, normal and heavy load operation conditions, basing on the theory of decompose and concert. Then three sub-models are coordinate by using heavy load loss time. Constraints conditions are solved by the use of Lagrange factor method and penalty function method, so the problem is transformed into problem of non-linear programming of no constraints. The conjugate gradient method is used in the solution and made a good convergence. In related power system calculation, a step by step method is used for tree shape system and Newton method is used for net shape system. The advantages are storage saved and efficient in calculation.This paper used PSO to console this model. PSO has a high efficiency and a simple model to solve optimize problem.