A Chaos Genetic Algorithm and its Application in Function Optimization
A chaos genetic algorithm (CGA) based on the ergodic property of chaos movement is proposed to solve the problems of optimization. Its basic principle is to load the chaos variables into the individuals of genetic algorithm (GA), and a small chaos disturbance is added to the child generation group, where, the disturbance amplitude is adjusted little by little as the search goes on. Furthermore, a type of improved adaptive genetic operators was presented to improve the convergence velocity and avoid the local convergence of GA. According to the concentrating extent of fitness for the population, the adaptive crossover probability and mutation probability were designed to adjust the crossover and mutation situations of the population. Finally, it was used in function optimization, theory analysis and simulation results had proved its property of fast convergence, and showed that it can avoid the premature convergence of GA in a great extent.
chaos genetic algorithm crossover mutation function optimization global convergence
Mingjie Chen Sheng Liu Changhong Wang
College of Automation, Harbin Engineering university, and Postdoctoral Research Station of Control S College of Automation, Harbin Engineering university China Postdoctoral Research Station of Control Science and Engineering, Harbin Institute of Technology Chi
国际会议
武汉
英文
2007-09-21（万方平台首次上网日期，不代表论文的发表时间）