Research On Dynamic Reactive Power Optimization In Distribution Network With Wind Power Generator

Reactive Power Optimization, as one effective means to improve voltage quality and reduce network loss, realizes the objective to reduce network loss and assure voltage quality by rational allocation of reactive-control equipments under each operation constraints. With the wide range of wind power generators (WPGs) connected into the distribution network, it will play an important roles in easing China’s power shortage and the goal of energy saving. However, the output of WPGs varies randomly with wind speed fluctuations, which causes more uncertainties to the conventional methods to reactive power optimization of distribution network. Therefore it is necessary to study the reactive power optimization in distribution network with WPGs connected in. Meanwhile, the traditional reactive power optimization is always operated under sectional load data, which not only neglect the dynamic nature of the load, but also not consider the number of operations of each control equipment. This hardly satisfies the demands of practical situation. Therefore, it is of great practical significance to study the dynamic reactive power optimization for distribution network.Adaptive Mutation Particle Swarm Optimization (AMPSO) and scene analysis are proposed and applied to solve the static reactive power optimization in distribution network with WPGs. Firstly, AMPSO with mutation operation is introduced, which can effectively enhance the ability of the particles to escape from the local optimal solution. Subsequently the approach of the power flow calculation with WPGs and its’random outputs description are presented in detail. The paper based on scene analysis using3deterministic scenarios effectively analyzes reactive power optimization problem in distribution network with WPGs. Example tests are conducted on static reactive power optimization in the United States PG&E69node system with WPGs in Matlab7.10. The results indicate that optimization programs based on scenario analysis have effective application in random output of WPGs, and it is easy to solve the problem by changing the uncertain output into3deterministic scenarios, which show the effectiveness of the modeling and solving difficulty for reactive power optimization.Least squares support vector machine is used to make short-term load forecasting in power system. The theoretical fundamental of the least squares support vector machine is analyzed and a brief introduction of short-term load characteristics is given. On this basis, considering the temperature, humidity, type of day and historical load data comprehensively as training samples. Then short-term load is forecasted based on LS-SVM. Finally, tests are conducted based on practical load data of a certain area in Sichuan, the forecasting results show that the LS-SVM is of great effectiveness in load forecasting.Dynamic reactive power optimization in distribution network with WPGs is solved based on the short-term load curve forecasted by least squares support vector machine. Firstly, the next day’s load forecasting natural curve is divided into24segments. Only the rated output scene is considered when WPGS connected in. Then AMPSO algorithm is used to make static reactive power optimization for each period to get the operation numbesr of each control device. Heuristic rules considering the product of the discrete control device value and the related-segment load change value is proposed to get the pre-dispatch schedule. Finally, considering the correlation of each control device during operation, dynamic adjustment of the control equipment is used to renew the dispatch schedule. The example test results show that the proposed method in solving dynamic reactive power optimization is of great adaptability and practical guiding significance.