Optimization On Suspension Structure Based On The Improved Genetic Algorithm

Vehicle handling and stability is one of the main criteria to measure the HyundaiMotor performance，which not only affect the manipulation of car’s driving, but alsodecided the safety of high-speed car. There are many vehicle handling and stabilityevaluations, so vehicle handling and stability optimization of automotive systems is alarge-scale multi-objective optimization problem, generally need the help ofprofessional multi-body system dynamics software to analyze and solve. Traditionaloptimization methods to deal with this type of optimization problems are often not easyconvergence and the results of the solving are local optimal solutions. GeneticAlgorithm is one of stochastic search algorithms, which has strong ability of globaloptimization. From home and abroad on the current situation of the genetic algorithmresearch, there are some problems in the Genetic Algorithm such as fitness evaluationstrategy, maintain the diversity of population strategies and algorithms convergence. Inthis paper, joint optimization of design and analysis platform which is based onimproved genetic algorithm is built to optimize vehicle suspension to improve vehiclehandling and stability.First, in view of the problems of the convergence of Genetic Algorithm, in thispaper the fixed point theory is used to improve the Genetic Algorithm. In the improvedgenetic algorithm, the K1subdivision is used to subdivide the solution space, and thencalculate the initial generation of population carrying the simplex, and label the simplexvertices, find out all the simplex, narrowing the scope of the solution space, speed upthe convergence speed of genetic algorithm; At the same time in the process of geneticalgorithm, according to the father and son prominence hybrid selection strategy toproduce individuals to avoid probability parameter selection problem.Second, the design variables and optimization objectives were determined. In thissection, first a car front suspension of4-degree-of-freedom model was established inADAMS/Car. Then, according to the relationship of the positional parameters of thevehicle suspension and vehicle handling and stability, the optimization of the objectivesand design variables were determined through the model simulation analysis; In order toimprove the efficiency of optimized, the ADAMS/Car and Isight software are united forsensitivity analysis of design variables to determine the main design variables. Finally, the joint optimization design and analysis platform based on improvedgenetic algorithm was built to optimize the vehicle suspension model. The platform isthe use of the powerful capabilities of process integration and secondary development ofmultidisciplinary integrated optimization design software Isight, improved GeneticAlgorithm, ADAMS/Car and MATLAB software together co-simulation analysis ofvehicle suspension model. Improved Genetic Algorithms is added to the optimizationalgorithm library of Isight software through the Isight secondary development ofsoftware module SDK and the JNI interface technology of the Java language. Theprocess of combination of ADAMS/Car and MATLAB software is use of the iSightsoftware integration components Simcode in the way of command scripts. With theestablished optimization analysis platform to optimize the vehicle suspension model, theoptimization results show that the joint optimization method feasibility andeffectiveness of improved genetic algorithm.