CN1889083A

CN1889083A - Car-human collision accident recurring optimizing method based on three-dimensional multi-rigid model

Info

Publication number: CN1889083A
Application number: CNA2006100291233A
Authority: CN
Inventors: 金先龙; 郭磊; 张晓云
Original assignee: Shanghai Jiao Tong University
Current assignee: Shanghai Jiao Tong University
Priority date: 2006-07-20
Filing date: 2006-07-20
Publication date: 2007-01-03

Abstract

A car-person collision accident reproduction optimization method based on a three-dimensional multi-rigid body model used in the field of traffic. The steps are: (1) establish a three-dimensional multi-rigid body mathematical model of the accident environment and the accident car and pedestrians; (2) determine the accident occurrence The position coordinates of the center of mass of the rear car stop and the position coordinates of the buttocks of the pedestrian; (3) Determine the optimization design variables, constraints and optimization objective function; (4) Call the optimization algorithm to optimize and solve the given objective function; (5) After many times After optimization cycle calculation, determine the convergence of the objective function and whether the optimal solution is obtained. The invention optimizes the collision process, shortens the reproduction time, reduces errors and uncertainties caused by human intervention, and can perform reproduction analysis on car-person collision accidents more accurately and efficiently.

Description

Optimization method for car-person collision accident reconstruction based on 3D multi-rigid body model

技术领域technical field

本发明涉及一种用于交通领域的碰撞事故再现优化方法，具体是一种基于三维多刚体模型的轿车-人碰撞事故再现优化方法。The invention relates to a collision accident reproduction optimization method used in the traffic field, in particular to a car-person collision accident reproduction optimization method based on a three-dimensional multi-rigid body model.

背景技术Background technique

随着我国汽车保有量的迅速增加，加上道路交通基础设施建设的相对滞后，导致交通碰撞事故发生频繁，其中汽车与行人发生的碰撞事故所占比例较高，约占总数的四分之一以上。由于实际碰撞事故发生的复杂性和现场遗留信息的多样性，加大了对事故进行再现分析的难度，不利于为交通管理部门正确鉴定事故提供快速准确的依据。With the rapid increase of the number of automobiles in our country and the relative lag in the construction of road traffic infrastructure, traffic collision accidents occur frequently, among which collision accidents between automobiles and pedestrians account for a relatively high proportion, accounting for about a quarter of the total above. Due to the complexity of actual collision accidents and the diversity of information left on the scene, it is more difficult to reproduce and analyze the accident, which is not conducive to providing a fast and accurate basis for the traffic management department to correctly identify the accident.

事故再现的关键在于根据事故现场的遗留信息来计算确定汽车碰撞速度，以此准确再现整个碰撞过程。对于汽车与行人碰撞事故，通常的再现研究方法有：动量/能量分析方法、变形/能量分析方法、抛撒物抛距分析方法等。动量/能量分析方法主要依赖汽车刹车印迹，将汽车和行人简化为二维集中质量模型，准确度不高，同时随着ABS技术在汽车上的普遍应用，很难得到清晰完整的刹车印迹；变形/能量分析方法考虑汽车车身碰撞变形，但是材料的非线性及弹塑性使得车身变形与碰撞速度之间的近似线性关系仅在一定范围内有效，同时利用有限元仿真计算的资源耗费巨大，效率低；抛撒物抛距分析方法在当抛出物遇到阻挡或者落点不确定时，会产生较大计算误差；可见，上述分析方法的局限性不利于对复杂的人车碰撞事故进行再现分析。The key to accident reproduction is to calculate and determine the vehicle collision speed according to the leftover information of the accident scene, so as to accurately reproduce the entire collision process. For car-pedestrian collision accidents, the usual reappearance research methods include: momentum/energy analysis method, deformation/energy analysis method, throwing object throwing distance analysis method, etc. The momentum/energy analysis method mainly relies on the brake marks of cars, and simplifies cars and pedestrians into a two-dimensional lumped mass model, which is not very accurate. At the same time, with the widespread application of ABS technology in cars, it is difficult to obtain clear and complete brake marks; deformation / The energy analysis method considers the collision deformation of the automobile body, but the nonlinearity and elastoplasticity of the material make the approximate linear relationship between the body deformation and the collision speed only valid within a certain range, and the resources consumed by the finite element simulation calculation are huge and the efficiency is low ; When the throwing object throwing distance analysis method encounters obstacles or the landing point is uncertain, large calculation errors will occur; it can be seen that the limitations of the above analysis methods are not conducive to the reproduction analysis of complex human-vehicle collision accidents.

随着近年来数字化假人的研究发展，Yang等在《International Journal ofCrash prevention and Injury Control》2000，Vol.2(2)：131-149上发表的“AHuman-Body 3D Mathematical Model for Simulation of Car-Pedestrian Impacts”，(“一种用于汽车对行人碰撞事故仿真研究的三维数值假人模型”，《国际碰撞预防和伤害控制杂志》，2000，Vol.2(2)：131-149)文中较早建立了三维行人数值假人模型进行人车碰撞事故的计算机模拟，具有较好的逼真度和准确性，但是其仿真再现过程需要人为根据单次计算结果不断计算调试，非程序自动优化完成，很大程度上依赖个人的经验，容易造成误差，增加了事故再现分析的时间，降低了效率。With the research and development of digital dummy in recent years, Yang et al published "A Human-Body 3D Mathematical Model for Simulation of Car- Pedestrian Impacts", ("A three-dimensional numerical dummy model for the simulation research of car-pedestrian collision accidents", "International Journal of Collision Prevention and Injury Control", 2000, Vol.2(2): 131-149). The 3D pedestrian numerical dummy model was established earlier for computer simulation of pedestrian-vehicle collision accidents, which has good fidelity and accuracy, but the simulation reproduction process requires continuous calculation and debugging based on a single calculation result, and non-program automatic optimization is completed. Largely relying on personal experience, it is easy to cause errors, increase the time for accident recurrence and analysis, and reduce efficiency.

因此，如何更加准确、高效率地对众多复杂的人车碰撞事故进行再现分析，是面临的一个迫切需要解决的问题。Therefore, how to more accurately and efficiently reproduce and analyze many complex collision accidents between people and vehicles is an urgent problem that needs to be solved.

发明内容Contents of the invention

针对目前事故再现存在的问题，本发明提出一种基于三维多刚体模型的轿车-人碰撞事故再现优化方法。本发明对碰撞过程进行优化，缩短了再现时间，减少了人为干预带来的误差和不确定性，可更加准确、高效率地对轿车-人碰撞事故进行再现分析。Aiming at the problems existing in the current accident reproduction, the present invention proposes a car-person collision accident reproduction optimization method based on a three-dimensional multi-rigid body model. The invention optimizes the collision process, shortens the reproduction time, reduces errors and uncertainties caused by human intervention, and can perform reproduction analysis on car-person collision accidents more accurately and efficiently.

本发明通过以下技术方案实现，具体步骤包括：The present invention is realized through the following technical solutions, and concrete steps include:

(1)建立事故发生环境及事故轿车、行人的三维多刚体数学模型；(1) Establish a three-dimensional multi-rigid body mathematical model of the accident environment and accident cars and pedestrians;

根据实际事故情况，建立包括路面、护栏、轿车、行人在内的三维多刚体碰撞数学模型，其运动方程为：According to the actual accident situation, a three-dimensional multi-rigid body collision mathematical model including road surface, guardrail, car and pedestrian is established, and its motion equation is:

$Σ Σ [[{δγ δγ}_{i i} (({m m}_{i i} {\overset{\cdot &Center Dot; \cdot &Center Dot;}{r r}}_{i i} - - {F f}_{i i})) + + {δπ δπ}_{i i} (({J J}_{i i} {\overset{\cdot &Center Dot;}{w w}}_{i i} + + {w w}_{i i} {\times \times J J}_{i i} {w w}_{i i} - - {T T}_{i i}))]] = = 00 - - - - - - ((i i = = 11,, . . . . . .,, n no))$

上式中m_i为质量，

为加速度，J_i为惯性矩，w_i为角速度，F_i为力，T_i为力矩，δγ_i为位移变量，δπ_i为初始方位变量。通过撞击试验可确定地面、护栏及轿车的碰撞接触刚度特性，依据人体生物力学试验数据设置行人模型的力学特性。In the above formula, m _i is mass,

is the acceleration, J _i is the moment of inertia, w _i is the angular velocity, F _i is the force, T _i is the moment, δγ _i is the displacement variable, and δπ _i is the initial orientation variable. Through the impact test, the impact contact stiffness characteristics of the ground, guardrail and car can be determined, and the mechanical characteristics of the pedestrian model can be set according to the human biomechanics test data.

(2)确定事故发生后轿车停止质心位置坐标、行人臀部落点位置坐标；(2) Determine the position coordinates of the car's center of mass when the car stops and the position of the pedestrian's buttock landing point after the accident;

根据事故现场环境，自定义一个固定的空间广义坐标系XYZ，然后可确定事故发生后轿车停止质心位置坐标C₁(x_c1，Y_c1，z_c1)、行人臀部落点位置坐标H₁(x_h1，y_h1，z_h1)。According to the accident scene environment, define a fixed space generalized coordinate system XYZ, and then determine the position coordinates C ₁ (x _c1 , Y _c1 , z _c1 ) of the center of mass of the car after the accident, and the coordinates H ₁ (x _h1 , y _h1 , z _h1 ).

(3)确定优化设计变量、约束条件和优化目标函数；(3) Determine the optimization design variables, constraints and optimization objective function;

事故中轿车碰撞速度和碰撞初始时刻位置为未知变量，故设定优化设计变量为轿车对行人的碰撞速度V，在坐标系XYZ中的碰撞初始时刻轿车质心位置坐标C₂(x_c2，y_c2，z_c2)、行人站立臀部位置坐标H₂(x_h2，y_h2，z_h2)。In the accident, the collision speed of the car and the position at the initial moment of collision are unknown variables, so the optimal design variable is set as the collision speed V of the car to the pedestrian, and the position coordinates of the center of mass of the car at the initial moment of collision in the XYZ coordinate system C ₂ (x _c2 , y _c2 , z _c2 ), the coordinates of the standing buttocks position of the pedestrian H ₂ (x _h2 , y _h2 , z _h2 ).

根据实际事故现场遗留信息，可初步估计轿车碰撞速度V的取值区间，同时坐标点C₂和H₂在轿车的行驶方向上相对于坐标点C₁和H₁存在一定的距离关系，通过此关系确定设计变量的约束条件。According to the information left over from the actual accident scene, the value interval of the collision velocity V of the car can be preliminarily estimated. At the same time, there is a certain distance between the coordinate points C ₂ and H ₂ relative to the coordinate points C ₁ and H ₁ in the direction of the car. Relationships determine the constraints on the design variables.

以坐标点C₁、H₁为优化目标点，确定优化目标函数：Taking the coordinate points C ₁ and H ₁ as the optimization target points, determine the optimization objective function:

$Y Y = = F f [[x x]] = = Min Min {Σ Σ}_{i i = = 00}^{m m - - 11} {β β}_{i i} f f (({x x}_{i i}))$

式中β_i为加权系数，一般可全取为1。In the formula, β _i is the weighting coefficient, which can be taken as 1 in general.

(4)调用优化算法对给定目标函数进行优化求解；(4) Call the optimization algorithm to optimize and solve the given objective function;

调用优化算法，在设计变量取值区间范围内进行搜索，经p次优化循环后计算得到碰撞速度V^p，轿车停止质心位置坐标C₁ ^p(x_c1 ^p，y_c1 ^p，z_c1 ^p)、行人臀部落点位置坐标H₁ ^p(x_h1 ^p，y_h1 ^p，z_h1 ^p)，使目标函数F^p[x]最小化，即：Call the optimization algorithm to search within the value range of the design variables. After p times of optimization cycles, the collision velocity V ^p , the position coordinates of the center of mass of the car at rest C ₁ ^p (x _c1 ^p , y _c1 ^p , z _c1 ^p ), The coordinates H ₁ ^p (x _h1 ^p , y _h1 ^p , z _h1 ^p ) of the pedestrian’s buttock landing point minimize the objective function F ^p [x], namely:

${MinF MinF}^{p p} [[x x]] = = Min Min (({[[{(({p p}_{c c 11}^{p p} - - {x x}_{c c 11}))}^{22} + + {(({y the y}_{c c 11}^{p p} - - {y the y}_{c c 11}))}^{22} + + {(({z z}_{c c 11}^{p p} - - {z z}_{c c 11}))}^{22}]]}^{11 / / 22} + + {[[(({x x}_{h h 11}^{p p} - - {x x}_{h h 11}))}^{22} + + {(({y the y}_{h h 11}^{p p} - - {y the y}_{h h 11}))}^{22} + + {(({z z}_{h h 11}^{p p} - - {z z}_{h h 11}))}^{22} {]]}^{11 / / 22}))$

(5)经多次优化循环计算后，判定目标函数收敛性，是否求得最优解。(5) After multiple optimization cycle calculations, determine the convergence of the objective function and whether the optimal solution is obtained.

经过p次优化循环计算，在满足约束条件的情况下，计算程序对每相邻两个优化循环得到的计算结果进行比较分析，根据设定的目标函数误差E^p判断是否收敛。目标函数误差计算公式为：After p times of optimization cycle calculations, if the constraint conditions are met, the calculation program compares and analyzes the calculation results obtained by every two adjacent optimization cycles, and judges whether the convergence is based on the set objective function error ^Ep . The calculation formula of the objective function error is:

${E E.}^{p p} = = | | \frac{{F f}^{P P} [[x x]] - - {F f}^{P P - - 11} [[x x]]}{{F f}^{P P - - 11} [[x x]]} | |$

E^p取值区间为[0，1]，随着优化次数p增加，E^p越小，表明目标函数误差序列呈递减趋势，两次相邻优化循环的计算结果越接近，优化结果越趋于收敛。当E^p值在设定的范围值内足够小，即可判断目标函数收敛，求得最优解。The value range of E ^p is [0, 1]. As the number of optimizations p increases, the smaller the E ^p is, it indicates that the error sequence of the objective function shows a decreasing trend. The closer the calculation results of two adjacent optimization cycles are, the closer the optimization results are to convergence. When the value of ^Ep is small enough within the set range, it can be judged that the objective function is converged and the optimal solution can be obtained.

本发明通过多刚体动力学方法和优化方法来进行事故再现，具有如下优势：1、该方法建立事故环境、轿车、行人的三维碰撞模型，碰撞过程更加直观，避免了将轿车和行人简化为二维平面集中质量模型带来的计算误差、可视化程度低等缺陷；2、利用多刚体动力学方法进行碰撞事故再现，效率更高，避免了有限元方法耗费计算资源大、再现计算时间长等缺点；3、确定设计变量和优化目标后，计算程序自动完成碰撞过程的优化再现，人为干预小，再现结果客观性程度高；4、该方法可模拟计算出行人各个部位在碰撞过程中所受伤害，便于与事故实际情况进行对比分析，同时能为法医鉴定提供参考和理论支持。The present invention reproduces the accident through a multi-rigid body dynamics method and an optimization method, and has the following advantages: 1. The method establishes a three-dimensional collision model of the accident environment, cars and pedestrians, and the collision process is more intuitive, avoiding simplifying the car and pedestrians into two 2. Using the multi-rigid body dynamics method to reproduce the collision accident is more efficient and avoids the shortcomings of the finite element method, which consumes a lot of computing resources and takes a long time to reproduce. ;3. After determining the design variables and optimization goals, the calculation program automatically completes the optimization reproduction of the collision process, with little human intervention and a high degree of objectivity in the reproduction results; 4. This method can simulate and calculate the injuries suffered by various parts of pedestrians during the collision process , which is convenient for comparative analysis with the actual situation of the accident, and can provide reference and theoretical support for forensic identification.

附图说明Description of drawings

图1为本发明实施例中的事故环境、轿车、行人三维多刚体模型Fig. 1 is three-dimensional multi-rigid body model of accident environment, car, pedestrian in the embodiment of the present invention

图2为本发明实施例中轿车-人碰撞事故现场图Fig. 2 is the scene diagram of car-person collision accident scene in the embodiment of the present invention

图3为本发明实施例效果图Fig. 3 is the effect diagram of the embodiment of the present invention

具体实施方式Detailed ways

本发明根据轿车与行人碰撞事故的特点和事故现场的遗留信息，选择轿车停止质心位置和行人臀部落点位置为优化目标，以目标函数误差值E^p为评价指标，它反映了第p次优化循环得到的计算结果是否趋于收敛。According to the characteristics of the car-pedestrian collision accident and the leftover information of the accident scene, the present invention selects the position of the center of mass of the car stop and the position of the buttocks of the pedestrian as the optimization target, and takes the error value E ^p of the objective function as the evaluation index, which reflects the pth optimization Whether the calculation result obtained by the loop tends to converge.

根据事故现场可定义一个固定的三维空间坐标系XYZ，然后确定事故发生后轿车停止质心位置坐标C₁(x_c1，y_c1，z_c1)、行人臀部落点位置坐标H₁(x_h1，y_h1，z_h1)。According to the accident scene, a fixed three-dimensional space coordinate system XYZ can be defined, and then the coordinates of the car’s center of mass C ₁ (x _c1 , y _c1 , z _c1 ) and the coordinates of the pedestrian’s buttock landing point H ₁ (x _h1 , y _h1 , z _h1 ).

以轿车对行人的碰撞速度V，碰撞初始时刻轿车质心位置坐标C₂(x_c2，y_c2，z_c2)、行人站立臀部位置坐标H₂(x_h2，y_h2，z_h2)为设计变量。通过计算机p次寻优计算，可以得到碰撞速度V^p，轿车停止质心位置坐标C₁ ^p(x_c1 ^p，y_c1 ^p，z_c1 ^p)、行人臀部落点位置坐标H₁ ^p(x_h1 ^p，y_h1 ^p，z_h1 ^p)，使目标函数F[x]最小，即：The car-pedestrian collision velocity V, car mass center position coordinates C ₂ (x _c2 , y _c2 , z _c2 ) at the initial moment of collision, and pedestrian standing hip position coordinates H ₂ (x _h2 , y _h2 , z _h2 ) are design variables. Through p-time optimization calculations by computer, the collision velocity V ^p , the position coordinates of the car’s center of mass C ₁ ^p (x _c1 ^p , y _c1 ^p , z _c1 ^p ), and the coordinates of the pedestrian’s buttock landing point H ₁ ^p (x _h1 ^p , y _h1 ^p , z _h1 ^p ), to minimize the objective function F[x], namely:

${MinF MinF}^{p p} [[x x]] = = Min Min (({[[{(({x x}_{c c 11}^{p p} - - {x x}_{c c 11}))}^{22} + + {(({y the y}_{c c 11}^{p p} - - {y the y}_{c c 11}))}^{22} + + {(({z z}_{c c 11}^{p p} - - {z z}_{c c 11}))}^{22}]]}^{11 / / 22} + + {[[(({x x}_{h h 11}^{p p} - - {x x}_{h h 11}))}^{22} + + {(({y the y}_{h h 11}^{p p} - - {y the y}_{h h 11}))}^{22} + + {(({z z}_{h h 11}^{p p} - - {z z}_{h h 11}))}^{22} {]]}^{11 / / 22}))$

F[x]越小，说明计算结果越接近真实事故情况，C₁ ^p、H₁ ^p分别与C₁、H₁接近于重合。The smaller F[x] is, the closer the calculation result is to the real accident situation, and C ₁ ^p , H ₁ ^p are close to overlap with C ₁ , H ₁ respectively.

由于数学优化方法存在一定的局限性，得到的优化结果与真实情况不可能完全一致，E^p反应优化结果是否趋于收敛，其值逐渐减小，说明优化结果呈收敛趋势，可最终求得F[x]最小值。Due to the limitations of mathematical optimization methods, the obtained optimization results cannot be completely consistent with the real situation. Whether the optimization results of E ^p response tend to converge, its value gradually decreases, indicating that the optimization results show a convergence trend, and finally F can be obtained. [x] Minimum value.

以下结合本发明的内容提供具体的实施例：Provide specific embodiment below in conjunction with content of the present invention:

以一起真实的轿车与行人碰撞事故为例，碰撞事故现场如图2所示。以轿车质心位置为坐标原点建立固定的空间坐标系XYZ(单位：m)，然后，可确定事故发生后轿车停止质心位置坐标(0，0，0)，行人臀部落点位置坐标(10.25，-4.66，-0.51)。Taking a real car-pedestrian collision accident as an example, the scene of the collision accident is shown in Figure 2. A fixed spatial coordinate system XYZ (unit: m) is established with the center of mass of the car as the coordinate origin. Then, the coordinates of the center of mass of the car after the accident occurs (0, 0, 0), and the coordinates of the pedestrian’s hip landing point (10.25, - 4.66, -0.51).

根据制动轨迹长度，可初步估计轿车行驶速度在30km/h～80km/h之间，碰撞初始时刻轿车与行人的位置在坐标系XYZ中可沿制动轨迹向X负方向移动0～40m距离之间。According to the length of the braking track, it can be preliminarily estimated that the driving speed of the car is between 30km/h and 80km/h, and the positions of the car and the pedestrian at the initial moment of collision can move 0 to 40m in the negative X direction along the braking track in the coordinate system XYZ between.

利用图1中建立的轿车与行人三维多刚体碰撞模型，以轿车对行人碰撞速度和碰撞时刻位置坐标为优化设计变量，轿车停止质心位置和行人臀部落点位置为优化目标，调用优化算法进行寻优计算。Using the three-dimensional multi-rigid body collision model between a car and a pedestrian established in Fig. 1, taking the collision speed of the car to the pedestrian and the position coordinates of the collision moment as the optimization design variables, the position of the center of mass of the car when it stops and the position of the pedestrian's buttocks as the optimization goals, the optimization algorithm is used to search Excellent computing.

通过多次优化循环计算，最后求得轿车与行人碰撞时刻车速为41km/h，轿车停止质心坐标(0.42，-0.13，0)，行人臀部落点坐标(10.01，-4.97，-0.56)，能很好重现事故情景，如图3所示。此时目标函数F[x]＝0.835，目标函数误差E^p＝0.0032，说明优化结果与真实事故接近，运用优化方法对碰撞事故进行了很好地再现分析。Through multiple optimization cycle calculations, the vehicle speed at the moment of collision between the car and the pedestrian is finally determined to be 41km/h, the coordinates of the center of mass of the car to stop (0.42, -0.13, 0), and the coordinates of the pedestrian's buttocks to fall (10.01, -4.97, -0.56). The accident scenario is well reproduced, as shown in Figure 3. At this time, the objective function F[x]=0.835, and the objective function error E ^p =0.0032, indicating that the optimization result is close to the real accident, and the collision accident is well reproduced and analyzed by using the optimization method.

Claims

1. A car-person collision accident reproduction optimization method based on a three-dimensional multi-rigid body model, characterized in that:

(1) Establish a three-dimensional multi-rigid body mathematical model of the accident environment and accident cars and pedestrians;

(2) Determine the position coordinates of the car's center of mass when the car stops and the position of the pedestrian's buttock landing point after the accident;

(3) Determine the optimization design variables, constraints and optimization objective function;

(4) Call the optimization algorithm to optimize and solve the given objective function;

(5) After multiple optimization cycle calculations, determine the convergence of the objective function and whether the optimal solution is obtained.

2. The car-person collision accident reproduction optimization method based on a three-dimensional multi-rigid body model according to claim 1, wherein said step (1) specifically refers to: for a real car-person collision accident, according to the accident In the actual situation, establish a three-dimensional multi-rigid body model of the accident environment including road surfaces and guardrails; establish a three-dimensional multi-rigid body model based on the geometric shape characteristics of the car; consider a pedestrian's height, age, and weight to establish a three-dimensional multi-rigid body model.

3. The car-person collision accident reproduction optimization method based on a three-dimensional multi-rigid body model according to claim 1, characterized in that, the step (2) specifically refers to: according to the accident scene environment, a custom space generalized coordinate system XYZ, and then determine the coordinates C ₁ (x _c1 , y _c1 , z _c1 ) of the center of mass of the car after the accident, and the coordinates H ₁ (x _h1 , y _h1 , z _h1 ) of the pedestrian's buttock landing point.

4. The car-person collision accident reproduction optimization method based on a three-dimensional multi-rigid body model according to claim 1, wherein said step (3) specifically refers to: setting the optimization design variable as the car-to-pedestrian collision Velocity V, coordinates C ₂ (x _c2 , y _c2 , z _c2 ) of the center of mass of the car at the initial moment of collision, and coordinates H ₂ (x _h2 , y _h2 , z _h2 ) of the pedestrian’s standing buttocks position, based on the estimated range of the car’s speed and the car’s , The distance relationship before and after the human collision is used as a constraint condition, and the optimization objective function is determined:

Y Y = = F f [[x x]] = = Min Min {Σ Σ}_{i i = = 00}^{m m - - 11} {β β}_{i i} f f (({x x}_{i i}))

In the formula, β _i is the weighting coefficient.

5. The car-person collision accident reproduction optimization method based on a three-dimensional multi-rigid body model according to claim 1, characterized in that, the step (4) specifically refers to: calling the optimization algorithm, within the value range of the design variable Search within , and after p times of optimization cycles, the collision velocity V ^p , the position coordinates of the car's center of mass C ₁ ^p (x _c1 ^p , y _c1 ^p , z _c1 ^p ), and the coordinates of the pedestrian's buttock landing point H ₁ p (x c1 ^p ) are obtained. _h1 ^p , y _h1 ^p , z _h1 ^p ), to minimize the objective function F ^p [x], namely:

Min Min {F f}^{p p} [[x x]] = = Min Min (({[[{(({x x}_{c c 11}^{p p} - - {x x}_{c c 11}))}^{22} + + {(({y the y}_{c c 11}^{p p} - - {y the y}_{c c 11}))}^{22} + + {(({z z}_{c c 11}^{p p} - - {z z}_{c c 11}))}^{22}]]}^{11 / / 22} + + {[[{(({x x}_{h h 11}^{p p} - - {x x}_{h h 11}))}^{22} + + {(({y the y}_{h h 11}^{p p} - - {y the y}_{h h 11}))}^{22} + + {(({z z}_{h h 11}^{p p} - - {z z}_{h h 11}))}^{22}]]}^{11 / / 22}))

.

6. The car-person collision accident reproduction optimization method based on a three-dimensional multi-rigid body model according to claim 1, characterized in that, the step (5) specifically refers to: after p times of optimization cycle calculations, when the constraints are met, In the case of , the calculation program compares and analyzes the calculation results obtained by every two adjacent optimization cycles, and judges whether the convergence is based on the set objective function error E ^p . The calculation formula of the objective function error is:

{E E.}^{p p} = = | | \frac{{F f}^{P P} [[x x]] - - {F f}^{P P - - 11} [[x x]]}{{F f}^{P P - - 11} [[x x]]} | |

The value range of E ^p is [0, 1]. As the number of optimizations p increases, the smaller the E ^p is, it indicates that the error sequence of the objective function shows a decreasing trend. The closer the calculation results of two adjacent optimization cycles are, the closer the optimization results are to convergence. When the value of ^Ep is small enough within the set range, it can be judged that the objective function is converged and the optimal solution can be obtained.