GATE：一种为昂贵的成本函数设计的遗传算法.pdfVIP

cost functions and Christophe Tribes Department of Mechanical Engineering, Ecole Polytechnique de Montreal, 2500,chemin de Polytechnique, Montreal (Quebec), H3T 1J4,Canada E-mail: stephane.dominique@polymtl.ca E-mail:jean-yves.trepanier@polymtl.ca E-mail: christophe.tribes@polymtl.ca *Corresponding author strengths and limitations of this algorithm. scale design problems; academic test cases. Optimisation, Vol.3,Nos.1/2,pp.5-29. design automation for gas turbine disks. multidisciplinary design optimisation. design optimisation problems and design under uncertainty. Copyright ◎ 2012 Inderscience Enterprises Ltd. 6 S.Dominique et al. 1 Introduction problems. GAs(Tsutsui, 2008; Eiben and Back, 1997). designs in a relatively low number of function evaluations. to better fit the specificities of each problem. literature. 2 Description of the GATE algorithm will be covered more thoroughly in subsequent subsections. 7 section. Figure 1 The GATE algorithm Define the pu anctus ro andam Define the end criteria Ern and om Initialize an active population Pof size N Initializethe total populationas P-P Fom cniough random couples to includc all individuds ofP. neva using the sane individal nwke 3|Fo each randomcouple.apply thennttionoparaor witha probability of a,. Apply crosovar opanaor othewiie Calculate. Evaluate ceach chiild locaed fatha than 6from evay previously evaluated solution Indude than in the totalpepulaion P Dacad ohe children 5|Constuc thenew active population Pusing theN