进化学习策略收敛性跟逃逸能力的研究探究资料.pdfVIP

31
6       Vol. 31, No. 6 2005  11 ACTA AUTOMATICA SINICA November, 2005    1)    ( !#$%#( )* 310027) (E-mail: xiaozhao9912@) + , -./0123456789:;= (? LELS) @01ABC56789:; = 7DEFGH7IJ:KLMNOPQRSTU NH7KLMWXYZ[ (? DELS) V GH7\]^ 0_^`:;7abcEdef9[VNH\]^7g9hijklm0_ @ ^7ngopq3Brstuvwx/yz{|7}~RU€`uv-./ DELS ‚ƒ „…7†‡ˆ‰Š‹o`ŒŽcq Ž•–8— ŽU˜™\xyz{| 8 ‘’“” ” ‚ ƒš›7œ†f9Š‹@š7}~žŸU c DELS 7ˆ‰Š‹„…o ¡¢£ 89:;=U23456UABC56ULELS DELS NOPQRST ¤¥¦§¨ TP273; TP30 Convergence and Escape Capacity Research of Evolution Learning Strategies YAN Ling JIANG Jing-Ping (College of Electrical Engineering, Zhejiang University, Hangzhou 310027) (E-mail: xiaozhao9912@) Abstract The Lamarckian evolution learning strategy (LELS) and Darwinian evolution learning strategy (DELS) are discussed in terms of their similarities and differences in learn- ing implementation. The former is based on inheritance of acquired character, i.e., both phenotype and genotype can be optimized through learning while the later only optimizes phenotype based on Darwinian selection. The convergence of ELS is proved using Markov chain theory. And we also theoretically demonstrate that DELS has stronger escaping capac- ity. These algorithms are applied to 8 standard test functions. Simulation results show that