The Research On Niching Particle Swarm Optimization In Dynamic Environments Based On Clustering

There are many optimization problems in the fields of industry, society, economy, and so on. As a new evolutional technology, swarm intelligence algorithm simulating some natural phenomenon, has made progresses on solving these optimization problems.Particle Swarm Optimization (PSO) algorithm is an evolutional computation technique based on swarm intelligence optimization algorithm, which was proposed by Kennedy and Eberhart in 1995 and inspired by the social behavior of birds blocking. Because its algorithm is simple and programmed easily, as one branch of swarm intelligence algorithm, PSO, has already been applicaed in many fieds successfully.The PSO algorithm is already successfully applied in the optimization of various static environments. However, many problems in real world are dynamic and changed stochastically over time, the current reasonable optimum is not certainly the optimum in the next time. It is essential to re-design the model of problem over time. It has realistic and active meanings to track changes of dynamic environments and search the optimum after a change.In order to track the movement of optimum with dynamic environments, the goal in dynamic environments is not only to detect the changes automatically but also to respond a variety of changes as timely as possible. From view of detection and response, a new improved PSO has been introduced and discussed in detail in this paper. The main researches are as follows:1. Introduced distribution evaluating strategy based on multi-sub-swarm, and using it to inspect the change of the environment. This strategy could reduce the overhead of algorithm and break the limitation of normal inspect methods that they were not able to work veraciously in variety environment.2. Raised the gist for response while environment changes. Analyzed the reason and necessary of raise these gist, then presented the relationship between the response gist, and relationship between response gist and resetting.3. Introduced a local optimum value crowding strategy which could keep the distribution and avoid the swarm converged to global optimum point.This method can avoid the decreasing of swarm variety, which is caused by converging of evolutional generations associate with their increasing.The strategy increased the variety of the swarms and enhanced the adaptabilities by dispersing the sub-swarm among different optimum point in search space. Finally, the development and research trend of PSO in more complicated environments are pointed out.