Reconstruction of pedestrian–vehicle accident using sequential linear programming optimizer

Abstract

The purpose of this paper is to present an optimization-based method for reconstructing pedestrian–vehicle accident and testing its performance. Using this method, the optimizer should automatically vary boundary conditions of multi-body dynamic simulation to generate the best match between simulation and real accident data. Considering the characteristic of pedestrian–vehicle collision, a corresponding optimization model is designed and sequential linear programming is used to deal with the non-linear objective function and the noisy constraints. By the reconstruction of two real-world pedestrian collisions, the performance of the method is tested. We compare the results with the global response surface of the objective function, which is constructed by a full-factorial experimental design. And the results are also evaluated in terms of injuries from the impact with the ground. The results show that this method is very effective in finding an optimum solution, which can not only reduce the number of cycles but also save manual operation.

Key words:

• Accident reconstruction
• pedestrian–vehicle collision
• approximation concepts
• simulation optimization

ACKNOWLEDGEMENTS

The authors gratefully acknowledge the Institute of Forensic Science of the Ministry of Justice of China for collecting real-world cases, the Shanghai Traffic & Police Office for support and cooperation, and the TNO Automotive China for providing free trial of MADYMO. This study is sponsored by the National High Technology Research and Development Program of China (No. 2007AA112234) and the National Natural Science Foundation of the People's Republic of China (No. 60174023).