Reactive Power Optimization Of Power System Based On Improved Genetic Algorithm

Reactive power optimization of power system can prevent accidents from expanding or causing and improve the voltage quality levels.Because of effectively reducing the system network loss and giving full play to economic benefits, reactive power optimization of power system became the main measure of making power system operation security,economic,stable and efficient.This is also the reason why the reactive power optimization has been studied for many years.In this dissertation,the research work is based on the developing characteristic of the reactive power optimization in modern power system.It puts the reactive power optimization problems as the research object,and chooses genetic algorithm as the main means.The model of the reactive power optimization mathematical model is built on the basis of mathematic model.It dealed with the state variable by the way of the 0,1 distribution variation so as to get the minimum network loss.In addition,the P-Q decomposition method can complete the reactive power flow calculation part rapidly in the case of high accuracy.At the same time,for the case that genetic algorithm can be easily trapped into local optimum in the application process of the reactive power optimization,this dissertation proposed a more effective reactive power optimization algorithm.The algorithm can contain all the possible solutions of the initial population according to the results of the initial power flow formulation.It used real integer code to make the same code string length and the same number of variables,while eliminating the cumbersome process of encoding and decoding to increase the availability.It designed fitness function of different phases and made optimization of the fitness at the beginning relatively larger to improve the efficiency of search algorithm.In the process of genetic manipulation,by improving the selection crossover and mutation,it increase the population diversity and produce more outstanding individuals.The combination of maximum genetic algebra and algebra of retaining the best individual make the optimization more scientific and reasonable.Finally,the dissertation provided the general program of reactive power optimization under genetic algorithm and improved genetic algorithm by the software platform of Matlab7.6,and calculated the bus system of the IEEE-6, IEEE-14 and IEEE-30. The results show that this optimization makes the speed of the optimization of reactive power faster and optimize the convergence better. Then it verifies the effectiveness of this algorithm,advantages and the real reference value.