Simulation And Optimization Of Occupant Restraint System Under Automotive Frontal Impact Condition

The produce and development of vehicle has brought convenience to people ’s travel, but also bring the security problem. The occupant restraint system, which can directly restraint and protect people from being injured during the vehicle crash, has been the main object in automotive passive safety research area. In the study of vehicle collision, researchers usually take use of numerical simulation method to do study, the method can greatly reduce the time, and improve efficiency. During the optimization design, the method based on agent model, which uses an approximate function to replace the complex simulation models, can get the desired optimal result in a short time. We choose an appropriate function according to the complexity of simulation model, the relationship between the input variables and output targets, and the same as choosing the sampling method and optimization method.In this paper, we chose vehicle occupant restraint system as the research object, and by use of using numerical simulation and surrogate mode to optimize several parameters which are sensitive to occupant injury, finally got the best parameter matching. At last, the optimized results were compared with the benchmark model to ver ify the result. The paper mainly includes the following aspects:(1)Establish the finite element model of occupant restraint system by LS-DYNA, then simulate the model according to US-NCAP evaluation requirements. Compare the result with actual vehicle test result to verify the validity of the model—benchmark model.(2)By use of agent model to optimize the benchmark model in software LS-OPT. Compare principles and characteristic of various approximate function, experimental design methods and optimization algorithm, then decide to select Radial Basis Function, Full fractorial, Genetic Algorithm(GA) and Adaptive Simulated Annealing(ASA) to make optimization calculation.(3)Compared the results generated by use of two optimization algorithms, analyzed the sensitivity of parameters, the relationship between the input parameters and the output response, the optimization effect, concluded that ASA is better than GA in aspect of vehicle optimization design. Compared the optimized dummy injury value by ASA with previous results, the optimization goal—WIC was reduced by 28.2%. The result showed that the optimized occupant restraint system parameters can effectively reduce the dummy injury value during frontal crash, thus can better protect the occupants.