The Parameter Estimation and Statistical Analysis of Constraint Optimization Penalty Function that Solved by Particle Swarm Optimization

摘要

Particle swarm optimization has been successfully applied to optimization, however it was not effective on all of the constraint functions. This paper has validated it through five measures and analyzed the result of test function. From the result we could find out which was the best one or which was the worst one, then we use four methods of Punishment strategy function and each of them was tested by four particle swarm optimization so that we could know which one is the best to be combined with particle swarm optimization, and it will be explained theoretically.