Multi-objective Optimizated Applications In The Safety Design Of Vehicle Collision

Vehicle safety design is becoming more and more important with thedevelopment of automobile industry and the increasing number of existing vehicles.Multi-objective optimization, which is a method for dealing with multiple objectives,can be well applied in the design of vehicle safety. In this paper, vehicle safety designis researched under several situations, including the frontal collision, side collisionand roll, in which the structural parameters are optimized by multi-objectiveoptimization. During the process of optimization, the impact of uncertainty on theresults and the difference between the simulation and experimental data areconsidered. The main contents of this paper are as follows:(1) A multi-objective genetic algorithm, which is combined with the uncertainty,is applied to optimize the structural optimization of frontal collision. In theoptimization process, utilize ellipsoid analysis approach to deal with objectivefunctions and constraint functions, in which the uncertainty’s effect on optimizationresults is taken into consideration. The interval is used to represent the affect ofuncertainty in the result.(2) A multi-objective genetic algorithm, which is combined with the uncertainty,is applied to optimize the structural optimization of side collison. Considering theuncertainty of material parameters, the objective functions and constraint functionsare solved by ellipsoid analysis method. The calculated results include influence ofthe uncertainty.(3) Utilize the multi-objective optimization technique to optimize the structuralparameters of car roll model, in which the properties of materials are obtained byexperiment. As the material parameters are fluctuated because of pre-processing andpost-aging treatment process, the parameters are obtained by combining experimentand simulation. First, manufacture an energy absorbing cylinder model with the carbody’s material. Then, an inverse technique is used to obtain the characteristicparameters by combining finite element simulation and test verification. Thecharacteristic parameters are applied to an optimization of rolling model.