Reactive Power Compensation Of The 10KV Distribution System On Ankang

Reactive power optimization of distribution system can not only advance the system's voltage quality, but also debase the active power loss effectively, and the aim that the system runs safely, economically and high-effectively can be achieved. This paper emphasizes on two matters in reactive power optimization: objective function and arithmetic of finding the best global individual. The author present the objective function considering the charge of active power loss and the investment of reactive compensation equipment at diverse load levels in order to get the economic balance between energy loss and compensator under the restrictions of voltage, compensator capability and power flow. Newton's method is used to calculate power flow. On the foundation of above, the conception of sensitivity analysis is introduced, and the nodes which provide higher sensitivity are selected in order to improve compensated effect and quicken the searching speed. Aiming at premature and slow convergent speed which happen easily in the cause of solving reactive power optimization which belongs to complicated Non-linear optimization, this paper improves the arithmetic, and self-fit cross rate and mutation rate are used to enhance the ability of looking for the global best individual. Therefore the speed and precision are boosted to a certain extentApplication program is developed according to the above model and arithmetic, it's function includes: analyzing power flow and sensitivity of anyone distribution system in which network topological structure and relevant parameters are told, compensation nodes are selected on demand; the charge proportion of each part in objective function can be adjusted in order to satisfy practical running demand; GA parameters can be adjusted such as group, evolution times, compensator numbers and compensator capability limits; finally data are showed such as compensator results and every branch power, branch power loss, branch power factor and every node's voltage value after compensating. The IEEE30 system and 10KV distribution system on AN KANG are employed to verify the effectiveness of the proposed model and algorithm, and the result shows that both the proposed model and algorithm are reasonable and feasible.