Convergence Analysis and Improvement of the Chicken Swarm Optimization Algorithm

Abstract

In this paper, the convergence analysis and the improvement of the chicken swarm optimization (CSO) algorithm are investigated. The stochastic process theory is employed to establish the Markov chain model for CSO whose state sequence is proved to be finite homogeneous Markov chain and some properties of the Markov chain are analyzed. According to the convergence criteria of the random search algorithms, the CSO algorithm is demonstrated to meet two convergence criteria, which ensures the global convergence. For the problem that the CSO algorithm is easy to fall into local optimum in solving high-dimensional problems, an improved CSO is proposed, in which the relevant parameters analysis and the verification of optimization capability are made by lots of test functions in high-dimensional case.

Keywords

• Chicken swarm optimization
• Markov chain
• state transition
• global convergence
• benchmark function