A parallel Multi-objective Genetic Algorithm On Cluster Computer
A parallel genetic algorithm to solve multiobjective optimization with discrete design variables on distributed shared memory is introduced. Real-value coding is used and one of PCs group is used to initialize group. The selection procedure as well as the crossover procedure is finished by different PC. Non-dominated Sorting Genetic Algorithm is used to compute the fitness of chromosome. A few examples formed by randomly are used to test the efficiency of the algorithm. The CPU time and computing efficiencies in the condition of different PCs number and different population size are given.
Genetic Algorithm Distributed share memory parallel computing Tread Marks
Lianshuan Shi Hui Liu
Computer Department, Tianjin University of Technology and Education,Tianjin 300222, China Department of vocational Education, Tianjin University of Technology and Education, Tianjin 300222,
国际会议
杭州
英文
47-49
2006-10-12（万方平台首次上网日期，不代表论文的发表时间）