Research And Application On Improved Genetic Algorithm Based On Gray Encoding And Real-number Encoding

Genetic algorithm is a random search method, which has very good ability for global searching. It is one of the optimal methods that is the most affecting and widely used currently. However, some deficiencies still exist in genetic algorithms, for example, subjectivity problem in probability parameters selection and prematurely problem more or less. Improvements are presented against the deficiencies motioned above in this paper. In the improved genetic algorithm base on gray encoding, the algorithm makes crossover and mutation organically combined, in which the all parent individuals are performed crossover to avoid the probability parameters selection, the high fitness offspring individuals instead of the low parent individuals to derive a new population. At the same time, the choice criterion of initial population is given which can get the highest fitness individuals. This algorithm has the advantages of fast convergence speed and small number of convergence. An other improve genetic algorithm base on real-number encoding ,this algorithm's idea come from schema theory, the merits this genetic operator lies in that in the early operation crossover and mutation act together, through mutation the size of population increase, while the population meet demand only do mutation operation. The algorithm can maintain the diversity and stability on the basis of population. It must find the optimal solution if there has enough time.Vehicle suspension system is a typical mutli-boby systems of precision space. And its performance is directly related to many vehicle capabilities, such as smoothly, the manipulation of stability, security and braking. It is greatly significant for the development of national vehicle industry to choose appropriate methods for optimization design of the vehicle suspension dynamics system parameters. Currently, the determinacy optimization methods are used to optimize the problems generally.This paper focuses on optimization design of complex suspension system parameter through above improve genetic algorithm. The result shows that the calculation quality and efficiency have greatly been improved compared with other methods currently.