The Effect on Structural Genetic Design of Binary Codin g Floating Coding and Noª²Uniform Mutation
Gao Feng, Hu Qiao, Wang Dejun
(School of Mechanical Engineering, Northeastern University)

Abstract: The effect on truss structural optimization of binary coding¡¢ float coding and nonuniform mutation is discussed. Best fitness and convergence rate of normal float coding are better than that of binary coding for solving minimum weight of truss structural problem. However, the average fitness is bad and approximately equal. In order to overcome this shortcoming, we adopt the noª²uniform mutation operator. The result shows that the best fitness and average fitness of noª²uniform mutation are improved. It also shows that the float coding is effective and convenient for truss structural design.
Keywords: genetic algorithms, structural genetic design, binary coding, float coding.