Particle swarm optimization (PSO) algorithm is a group nuclear intelligent algorithm developed in recent years which based on search strategy. This algorithm is based on the idea of every particle in the group can receive effective information from past experience as well as the experience of other particles to amend its own search path.This algorithm is widespread concerned because of its simplicity and effectiveness. However, the research of PSO algorithm is not perfect; there are few discussions of the following aspect. Firstly, the research of determining the relationship of the fitness among objectives based on multi-objective optimization, especially in its high-dimensional multi-objective optimization is almost never seen. Secondly, many factors changed with time in practical applications, how to polish the traditional PSO algorithm to make it applicable to the dynamic optimization problem is also never seen.In this dissertation, some relevant improvements for particle swarm optimization algorithm are deduced for multi-objective and dynamic optimization problems. An ideal gray particle swarm optimization (IGPSO) algorithm suitable for solving high-dimensional multi-objective optimization and a presuppose best particle swarm optimization (PBPSO) algorithm are proposed.Dynamic multi-objective optimization problems in the project are often the most common and most complex, at the end of this dissertation, an ideal gray particle swarm optimization (IGPSOS) algorithm based on the combination of the ideal gray particle swarm optimization and a variety of group thinking is proposed.These three algorithms presented by this dissertation perform well for their specific issues through the relevant experimental test. Ideal gray particle swarm optimization (IGPSO) algorithm has advantages of analyzing situational change and the curve's geometry shape similarity of un-inferior data curve and ideal solution data curve as well as showing the position relationship of the curve between un-inferior data curve and ideal solution data. Presuppose best particle swarm optimization (PBPSO) algorithm is able to forecast the range of scope in circumstance of little environment change, so as to accelerate the relevant searching speed. (PBPSO) algorithm also performs well in circumstance of enormous environment change. Ideal gray particle swarm optimization (IGPSOS) algorithm performs well through using test function simulation; which can solve the problem of dynamic multi-objective optimization well. |