Reverse reconstruction of motorcycle-car accident based on response surface model and NSGA-II algorithm. Issue 1 (2nd January 2022)

• Record Type: Journal Article
• Authors: Wang, Qian; Lou, Yunfeng; Li, Tong; Jin, Xianlong; Kong, Lingshuang; Hou, Xinyi
• Is Part Of: International journal of crashworthiness. Volume 27:Issue 1(2022)
• Page Start: 107
• Page End: 116
• Publication Date: 2022-01-02
• Subjects: Motorcycle traffic accident -- reverse reconstruction -- initial collision conditions -- response surface model -- NSGA-II algorithm; Automobiles -- Crashworthiness -- Periodicals; Airplanes -- Crashworthiness -- Periodicals; 629.2042
• DOI: 10.1080/13588265.2020.1774482
• Languages: English
• ISSNs: 1358-8265

Abstract

Abstract: In order to obtain the initial condition of the traffic accident and reconstruct the collision process of motorcycle and car in reverse, this paper proposes a multi-objective optimisation algorithm based on the combination of NSGA-II and surrogate model. Aiming at a collision accident between a car and motorcycle, a simulation model of the human-vehicle-road coupling system is established. First, select the optimisation variables, optimisation objectives and constraints, define the variable range, and through the Opt-LHD to sample the optimisation variables. Then, obtain and fit the sample data, and take the sample data as the initial conditions into the established motorcycle traffic accident simulation model. Next, obtain the simulation data of the optimisation variables, and establish the response surface model of the optimisation objectives in MATLAB. Finally, using NSGA-II algorithm, the individuals in the population are solved based on the response surface model in turn, the Pareto solution set is obtained, and a group of satisfactory solutions is taken back to the simulation model for verification of the optimisation model, so as to prove the correctness of the satisfactory solution, and then the initial conditions of the collision between the motorcycle and the car are obtained.