The Research Of Multi-objective Optimization Methods In Economic Dispatch Of Power System And Reactive Power Optimization

The multi-objective optimization is a method to solve practical engineering problem. Most issues in the present world relate to multiple objectives, and these objectives are not independent, they are related to each other. Many problems in power system refer to multiple objectives, the application of multi-objective optimization in power system can result in multiple solutions and schemes from which the operation staff could choose according to the actual situation. After introducing the meaning of economic dispatch of power system and reactive power optimization, several multi-objective optimization methods which are used in economic power dispatch and reactive power optimization are studied in this paper, the structure of this paper is as follows:Firstly, after the application of Paralleled Particle Swarm Optimization (PPSO) in reactive power optimization is elaborated, the Multi-objective Comprehensive Learning Particle Swarm Optimization method is put forward by combining the Multi-objective Comprehensive Learning scheme with the PSO algorithm. After that, this proposed method is applied into the economic dispatch of power system. In this method the particle can not only learn from its own previous best solution (pbest), but also learn from all other particle's previous best solutions, therefore, the diversity of the particle swarm can be guaranteed, the possibility of getting premature can be reduced, and global searching ability will be promoted.Secondly, after expatiating on the application of Interior Point Method (IPM) in economic dispatch of power system, the IPM is modified by adding the Satisfaction-Maximizing Decision Approach into it, and the modified method is applied into the economic dispatch of power system. In this algorithm, the objective functions of fuel cost and emission are optimized respectively, after that, the solutions of those functions are regarded as the upper and lower bound of these contradictory objective functions. Bi-objective fuzzy function can refer to these upper and lower bound to create a set of non-dominated solutions by using the fuzzy-satisfaction maximizing scheme.Thirdly, after analyzing the application of Simulated Annealing (SA) in multi-objective economic dispatch, the dual-layer SA is proposed in this paper. This modified algorithm can be divided into analogy layer and decision layer. Decision layer proposes the ideal point and judge the solutions obtained by analogy layer. In the analogy layer, new feasible points can be gained by alternately refreshingλand simulated annealing method. These processes carry out alternately till the optimal solution that meets the condition is found.Focusing on the economic dispatch of power system and reactive power optimization, three kinds of commonly used multi-objective optimization methods are elaborated in this paper, and three novel optimization methods are proposed by the modification of traditional ones. The methods proposed in this paper have been tested in the IEEE30 system. Comparing the methods presented in this paper with the others, the test result showed that the former ones are of better global searching ability and greater efficiency.