Research On Reactive Power Optimization Of Power System Based On Hybrid Intelligent Algorithm

Rational distribution of reactive power in power system is directly related to system power quality, while power quality and reduction degree of the system network loss are the precondition of safe, stable and economical system operation. Reactive power optimization of power system is not only the important means to realize economical operation of power system, but the basic condition for good power quality and safe operation of power system. Therefore, it is of great significance in theory research and practical application. Some of the traditional optimization algorithms, such as linear programming and nonlinear programming and so on, must rely on the accurate mathematical model, and they also usually have other requirements for the object function of solved problem. What’s more, while reactive power optimization problems which contain discrete variables are dealed with by some of the traditional optimization algorithms, a biggish error would be involved in their results. In recent years, the application of artificial intelligence methods in the field of reactive power optimization of power system has become a hot research topic.Firstly, this dissertation introduces the research background and significance of reactive power optimization of power system, then briefly describes the basic characteristics of the traditional methods and the modern artificial intelligence algorithms, which are applied to reactive power optimization of power system. On the basis of lucubrating reactive power optimization of power system, this dissertation analyzes the reactive power optimization model of power system and the reactive power control equipment used commonly in power system according to the characteristics and requirements of power system, and focuses on the mathematical model and realization methods of the power flow calculation. With the in-deepth research on the mathematical models and algorithms of reactive power optimization, this dissertation builds the mathematical model of reactive power optimization respectively based on genetic algorithm, ant colony algorithm or hybrid intelligent algorithm which focuses on the fusion of pheromone updating strategy and crossover operation, and the corresponding object function is confirmed, too. Considering the situation of the voltage cross-border, reactive power cross-border and transformer voltage ratio cross-border in the.mathematical model, focusing on voltage stability, the cross-border penalty factors and penalty functions are adopted to put the constraint conditions into the object function, and the constrained problem will be converted into the unconstrained problem.In order to verify the feasibility and superiority of the three algorithms above, the IEEE30-bus standard testing system is chosen as the optimized object, and the corresponding programs based on them are made by MATLAB language to realize reactive power optimization calculation. And the optimization results between the algorithms are compared. The results show that the hybrid intelligence algorithm proposed in this dissertation is visibly better than the other two algorithms in terms of the searching speed and convergence property, and it meets the basic requirements of reactive power optimization, meanwhile the feasibility and validity of the proposed algorithm and model are proved.