一种新型的自适应混沌遗传算法.pdfVIP

4 Vol. 34 N o. 4 2006 4 ACTA ELECTRON ICA SIN ICA A pr. 2006 1 2 1 1 袁晓辉 , 袁艳斌 , 王 乘 , 张勇传 ( 1, 430074; 2, 430070) : , . , , , , , ,. , , . : ; ; ; : TM 612 : A : ( 2006) A N ovel Selfadap tiv e Chaotic G en etic A lgorithm 1 2 1 1 UAN X iaohui, UAN anbin, WANG Cheng , ZHANG ongchuan ( 1College of Hydropow er Engineering, HuazhongU niversity of Science Technology,Wuhan,Hubei 430074, China; 2. College ofResource EnvironmentalEngineering,Wuhan University of Technology,Wuhan, Hubei 430070, China) A bstract: This paperpresents a new realvalue encoding selfadaptive chaotic genetic algorithm to solve optmi i zation problem based on the analysisfor shortcom ing of standard binaryencoding genetic algorithm. It isused the entro py based on inform ation theory to initialize populationw ith better distribution and designed a crossover operator in light of probability distribution function and a selfadaptive chaoticmutation operator combined chaotic dynam ic character w ith artificial neuralnewt ork theory, w hichm aintains population diversity to preventand overcome prem ature phenom e na in the evolutionary process. This algorithm is easy to mi plem entw ith the smi ple operation. Several typicalbenchmark function numerical expermi ents demonstrate that it is mi proved on the solution precision and increased convergence speed. The proposed m ethod provides an effective new m ethod to solve the function optmi ization problem s. K ey w ords: genetic algorithm; chaos; prem ature convergence; entropy