The Research And Application Of Dynamic Multi-objective Optimization Based On Particle Swarm Optimization

In the field of intelligent algorithm, except for evolutionary algorithm based on the idea of survival of the fittest of Darwin has accept success to some degrees in solving optimization problem, PSO which is one of swarm intelligence also has got success. Particle Swarm Optimization is one of the relatively new optimization techniques, whose main idea is swarm intelligence effect, and it’s based on simulating fish schooling and birds flocking. It is easy to understand in concept, convenient application and rapid convergence. And compared with other evolutionary algorithms, it can get better results with fewer individuals search, low storage capacity. It aroused the interest of many researchers and gradually become an important branch of evolutionary algorithm, since particle swarm optimization algorithm is proposed by Kennedy and Eberhart in1995.Evolutionary algorithms have achieved good performance in solving the static multi-objective optimization problems. In real life, many problems not only have multiple targets, but also relate to time, so the problems which relate to time and have more than one type of objective functions to be optimized simultaneously are called dynamic multi-objective optimization problems. Since dynamic multi-objective optimization problems have many conflicting and incommensurable objectives depended on time (environment), and the pareto optimal solutions change with time, so the optimization is relatively difficult. It brings new challenges to the particle swarm algorithm and evolutionary algorithm.The main work contains several aspects as follows:The first part gives a brief introduction of optimization problems, mathematical models, traditional evolutionary algorithms used to solve the optimization problem and swarm intelligence algorithms based on evolutionary algorithms, highlights the research of solving multi-objective optimization problem and dynamic multi-objective optimization problem, summarizes strategies which were used to improve the performance of the algorithm.The second part describes the particle swarm optimization algorithm, analyses the Particle Swarm Optimization and introduces several kinds of commonly used method of improved particle swarm optimization algorithm firstly. Then it focuses on the idea of dynamic multi-swarm particle swarm optimization algorithm, analysis the advantages of dynamic multi-swarm particle swarm optimization algorithm, which shows the effectiveness of the dynamic multi-swarm particle swarm optimization algorithm.The third part, first of all, in order to further confirm the validity of the dynamic multi-swarm particle swarm optimization algorithm used to solve multi-objective problems, the practical application of the environmental/economic dispatch is simulated in the standard IEEE30-bus6-generator test system, which further confirms the validity of the dynamic multi-swarm particle swarm optimization algorithm to solve multi-objective optimization problem.The forth part gives a mathematical description about the mode of changes in dynamic environment and the types of changes of a dynamic multi-objective problem, analyses the performance evaluation criteria of the algorithm, then carries on the simulation using dynamic multi-swarm particle swarm optimization algorithm which is based on Gaussian variation on the relevant dynamic multi-objective test function, verified algorithm can track the transformation of the Pareto frontier fast and Pareto optimal solution has a better distribution.Finally, it carries on the full text summary, explains the deficiencies of this article, and looks forward to the future of the development and application of particle swarm optimization algorithm.