Optimization of a Child Restraint System by Using a Particle Swarm Algorithm

Abstract

Child restraint system (CRS) is a system in automotive vehicles for the protection of child occupants in traffic accidents. Design of appropriate CRS has been one of the major subjects for both the research community and the automotive industry. In this paper, a CRS, which includes a child booster and an adult seatbelt with load limiting function, is optimized for a ten-year child dummy. The model is built and simulated using MADYMO. Several key parameters of the system are optimized to minimize the injury to child passengers under the crash test circumstance in accordance with the ECE Regulation 44 by using a recently emerged optimization scheme, particle swarm algorithm. In order to validate this optimization approach, another optimization method, AutoDOE, a built-in subroutine of MADYMO, is also utilized for comparison. The results indicate that the particle swarm algorithm has certain advantages over the AutoDOE method in terms of the solution quality. Moreover, regarding the computational efficiency, for this particular problem the particle swarm algorithm outperforms AutoDOE.