CN101173856B - Reconstruction Method of Automobile Collision Accident Based on Photogrammetry and Body Outline Deformation - Google Patents

Abstract

A method for reproducing automobile collision accidents based on photogrammetry and deformation of the outer contour of the vehicle body used in the field of transportation. The three-dimensional numerical model of the same model and intact vehicle outer contour; analyze and compare the benchmark model and the deformation numerical model, and obtain the deformation amount and deformation angle of the feature points; establish the local finite element corresponding to the accident scene environment and the vehicle deformation parts involved in the accident Model; through multiple iterative optimization calculations, under the conditions of set speed and collision angle, the numerical simulation results of the deformation of feature points are consistent with the real collision measurement results, so as to accurately determine the vehicle speed and collision angle at the time of the accident. The invention can realize accident reproduction under the condition of accident scene information, effectively improves the efficiency of accident reproduction, and has more objectivity and intuitiveness.

Description

Reconstruction Method of Automobile Collision Accident Based on Photogrammetry and Body Outline Deformation

technical field

The invention relates to a method used in the technical field of transportation, in particular to a method for reproducing an automobile collision accident based on photogrammetry and deformation of the outer contour of a vehicle body.

Background technique

In order to reproduce the analysis of automobile collision accidents, usually from multiple angles such as energy and momentum, according to the characteristic parameters of the collision stage (such as vehicle deformation characteristics, rebound coefficient, contact surface friction coefficient, collision center, etc.) Image, use lumped parameter model method and multi-rigid body system dynamics method to analyze the accident. At present, the commonly used accident analysis methods in the world include impulse/momentum analysis method and deformation/energy analysis method. In addition, throwing distance analysis methods, geometric and time analysis methods, and centrifugal force analysis methods are also applied under specific conditions.

After searching the literature of the prior art, it was found that the Chinese patent application number: CN200510027720.8, this patent relates to a car collision accident reproduction method based on the three-dimensional deformation of the key points of the car body, the method first uses the three-coordinate measuring instrument to deform the inner contour of the car body The key points of the accident are physically measured, and then the finite element model of the accident scene and the accident vehicle is established. Finally, through multiple iterative calculations, the numerical simulation results of the deformation of the key points are consistent with the real collision measurement results, so as to reproduce the accident. However, on the one hand, the three-coordinate measuring instrument is relatively expensive, on the other hand, it requires professionals to operate, and it takes a long time. In addition, the finite element model of the whole vehicle used in this method makes the modeling and simulation calculation process extremely consuming. , the efficiency of accident reproduction is greatly reduced, which does not meet the requirements for rapid handling of traffic accidents.

Lu Guangquan and Li Yibing published "Traffic Accident Photogrammetry Technology and Research Progress Based on Ordinary Digital Cameras" in "Journal of Transportation Engineering and Information Science" 2005.Vol3(9): 63~67. Application and research status of photogrammetry technology in the field of traffic accidents. The article points out that 3D photogrammetry technology has been applied in the field of accident reconstruction since the 1990s, and is currently mainly used in 3D modeling of traffic accident scenes and environments, and 3D modeling of vehicle and body deformation is still in the research stage.

Contents of the invention

In view of the above-mentioned deficiencies and defects in the prior art, the present invention proposes a method for reproducing automobile collision accidents based on photogrammetry and deformation of the outer contour of the vehicle body, so as to replace the method for measuring the deformation of the vehicle body using a laser rangefinder and a three-coordinate measuring instrument , as well as the accident reproduction method of the vehicle finite element, and can solve the reproduction analysis of traffic accidents in the case of no brake marks or brake marks are not clear.

The present invention is achieved through the following technical solutions, and the present invention comprises the following steps:

The first step is to conduct photogrammetry of the deformed vehicle involved in the accident and the same model and intact vehicle;

The photogrammetry refers to: place four identical photogrammetry calibration objects evenly around the vehicle to form a rectangular area, take the rectangular area as the center of photography, and use a digital camera to take a four-point view of the vehicle site in a clockwise direction , the adjacent shooting orientations form an angle of approximately 90°, and ensure that each photo contains at least three calibration objects, and measure and record the relative positions between the four calibration objects, so as to realize the photogrammetry calibration of the vehicle site.

The second step is to establish a three-dimensional numerical model of the outer contour of the deformed part of the accident vehicle and the outer contour of the same model and intact vehicle;

Import the captured photos of the deformed vehicle into the computer, select a certain amount of points on the deformed contour of the vehicle as feature points, and obtain their pixel coordinate values, and establish the image space coordinate system and the object space coordinate system (coordinates in the actual space) according to the DLT method. ), the object space coordinates of these points are iteratively obtained by the least square method, and then the related points are connected to establish a three-dimensional numerical model of the outer contour of the deformed part of the vehicle. These feature points can describe the deformation profile of the vehicle, and are usually selected near the frame of the car, because these places are generally more rigid and are the main energy-absorbing areas. For example, for frontal and rear-end collisions, they can be selected on the front and rear bumpers , For side impacts, it can be selected on the bottom frame of the door.

Use the same method to select the feature points of the outer contour of the intact vehicle, establish the outer contour model of the vehicle and use it as the benchmark numerical model. For frontal collision or oblique collision, the contour features of the front part of the body should be refined; for side collision, the side parts of the body should be described in detail, and other places can be simplified appropriately.

The third step is to analyze and compare the reference model and the deformation model, and use the vector method to obtain the deformation amount at the feature point and the inclination angle of the deformation contour;

Match the undeformed features (such as point features or line features) on the deformed car body with the corresponding features of the undeformed vehicle of the same model, and then import the benchmark numerical model into the model of the deformed vehicle, so that the deformed part is intact The relative position in the whole vehicle can be determined, and then the characteristic points of the deformed part and the coordinates of the corresponding points in the intact model are picked, and the deformation amount at each characteristic point is calculated, and then the inclination angle of the deformed contour is calculated.

In the matched model, select two feature points a and b that can represent the deformation trend, assuming that their coordinates are (x _a , y _a , z _a ), (x _b , y _b , z _b ), corresponding to The coordinates of the corresponding points a′ and b′ in the reference model before deformation are (x′ _a , y′ _a , z′ _a ), (x′ _b , y′ _b , z′ _b ) respectively, then it can be considered that The deformations of feature points a and b in the accident are respectively:

$\left\{\begin{array}{c}{\mathrm{<span\; class="notranslate">\; l</span>}}_{\mathrm{<span\; class="notranslate">\; a</span>}}<span\; class="notranslate">\; =</span>\sqrt{{<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; a</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; a</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; +</span>{<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; a</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; a</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; +</span>{<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; a</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; a</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}\\ {\mathrm{<span\; class="notranslate">\; l</span>}}_{\mathrm{<span\; class="notranslate">\; b</span>}}<span\; class="notranslate">\; =</span>\sqrt{{<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; b</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; b</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; +</span>{<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; b</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; b</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; +</span>{<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; b</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; b</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}\end{array}\right.$

Then the inclination angle of the deformed contour can be considered as:

$\mathrm{<span\; class="notranslate">\; \&alpha;</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; arctan</span>}\frac{<span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; l</span>}}_{\mathrm{<span\; class="notranslate">\; a</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; l</span>}}_{\mathrm{<span\; class="notranslate">\; b</span>}}<span\; class="notranslate">\; |</span>}{\sqrt{{<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; a</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; b</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; +</span>{<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; a</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{\mathrm{<span\; class="notranslate">\; b</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; +</span>{<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; a</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; z</span>}}_{\mathrm{<span\; class="notranslate">\; b</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}}$

The fourth step is to establish the finite element model of the accident scene environment and the deformation parts of the vehicles involved in the accident;

Based on the body design drawings or the 3D numerical model established in the second step, establish the corresponding finite element model of the deformed part of the vehicle. The mesh at the deformed part needs to be refined, and the unit length is not greater than 10mm. Other parts can be appropriately Simplified, the area far away from the deformation can be directly replaced by a rigid body.

In the fifth step, through multiple iterative optimization calculations, under certain speed and collision angle conditions, the numerical simulation results of the deformation of the feature points are consistent with the real collision measurement results, so as to accurately determine the speed of the vehicle (single vehicle or two vehicles) at the time of the accident and the angle of impact.

Firstly, a comprehensive analysis is carried out based on the evidence obtained from the accident scene investigation, and the range of the vehicle speed and collision angle is initially determined; The deformation at the corresponding feature points in a group of finite element models can be obtained through calculation; then the photogrammetry results and numerical simulation results are compared according to the deformation fit formula, when the numerical simulation results are closest to the deformation of the feature points in the real accident results , it is considered that the collision speed (single vehicle or two vehicles) and collision angle corresponding to the simulation result are the real collision speed and collision angle.

Deformation fit degree M formula, M=1-(∑(x _iP -x _iA ) ² /(∑x _iP x _iA ))

Among them, x _iP is the actual deformation amount at the feature point obtained by photogrammetry, and x _iA is the deformation amount at the feature point obtained through numerical simulation.

The larger the M value, the closer the accident reproduction process is to the real situation. Taking the deformation coincidence degree M as the objective function, the speed of the vehicle (single vehicle or two vehicles) before the collision and the mutual collision angle are used as optimization variables, and the value of M is maximized by selecting an appropriate optimization method for repeated iterative calculations.

Due to the limitations of numerical simulation, it is impossible to completely match the simulation results with the real collision results. The value of M can qualitatively reflect the gap between the accident reproduction results and the real results, and at the same time reflect the impact of speed changes and angle changes. Influence of M-values on trends. For the definition of M value, when the numerical simulation result is close to the real result, the M value tends to "1".

The variation interval of the M value is [m, 1], where m can be positive or negative, and the specific value is determined by the closeness between the simulation result and the real collision accident.

The present invention has the following advantages: the accident reproduction method based on the deformation of the vehicle body can solve the problems that cannot be solved by common accident reproduction methods in the absence of accident site information such as brake marks; Model, and then measure the deformation at the feature points, the method is simple and intuitive, with high accuracy and low cost; and the measured information such as the inclination angle of the deformation contour can provide important initial conditions for accident reproduction, thereby narrowing the search range. Improve the efficiency of accident reproduction; the finite element method can be used to conduct three-dimensional dynamic inspection of the main deformation parts of the car body, which can intuitively reproduce the deformation process and provide reference for the automobile design department; the local finite element method can reduce the cost of vehicle modeling and simulation calculations Reduce the workload and improve the efficiency of accident reproduction.

Description of drawings

Fig. 1 is the flowchart of the accident reproduction method of the present invention;

Fig. 2 is a schematic diagram of photogrammetry of a deformed vehicle in an embodiment of the present invention;

Fig. 3a is the outline diagram of the whole vehicle of the reference vehicle in the embodiment of the present invention;

Fig. 3b is a three-dimensional numerical model diagram of the outer contour of the reference vehicle in the embodiment of the present invention;

Fig. 4a is the outline drawing of the deformed part of car A in the accident in the embodiment of the present invention;

Fig. 4b is a three-dimensional numerical model diagram of the outer contour of the deformed part of car A in the accident in the embodiment of the present invention;

Figure 5a is a matching diagram of the deformed vehicle benchmark model and the deformed model in the embodiment of the present invention;

Fig. 5b is a measurement diagram of the deformation amount of the deformed vehicle in the embodiment of the present invention.

Detailed ways

Below in conjunction with accompanying drawing, embodiment of the present invention is described in detail: present embodiment implements under the premise of technical scheme of the present invention, has provided detailed embodiment and concrete operation process, as shown in Figure 1, but the present invention The scope of protection is not limited to the following examples.

Taking a real typical car-vehicle oblique collision accident as an example, here we only take the modeling and measurement of car A in the accident and the method of accident reproduction as examples for illustration.

Firstly, put the deformed vehicle into a rectangular area surrounded by four photogrammetric calibration objects with known relative positions. With the rectangular area as the center, the Panasonic-LX2 10-megapixel ordinary digital camera was used to make a circular image of the vehicle. shooting, as shown in Figure 2. In order to achieve a better photogrammetry effect, the camera intersection angle between two adjacent shooting orientations is close to 90°, and it is guaranteed that each photo contains at least three calibration objects.

Then import the obtained photos into the computer, and establish the direct linear transformation relationship between the image space coordinates and the actual object space coordinates according to the linear transformation relationship between the object space coordinate system and the image coordinate system:

$\left\{\begin{array}{c}{\mathrm{<span\; class="notranslate">\; I</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; +</span>\frac{{\mathrm{<span\; class="notranslate">\; l</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; l</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}\mathrm{<span\; class="notranslate">\; Y</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; l</span>}}_{\mathrm{<span\; class="notranslate">\; 3</span>}}\mathrm{<span\; class="notranslate">\; Z</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; l</span>}}_{\mathrm{<span\; class="notranslate">\; 4</span>}}}{{\mathrm{<span\; class="notranslate">\; l</span>}}_{\mathrm{<span\; class="notranslate">\; 9</span>}}\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; l</span>}}_{\mathrm{<span\; class="notranslate">\; 10</span>}}\mathrm{<span\; class="notranslate">\; Y</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; l</span>}}_{\mathrm{<span\; class="notranslate">\; 11</span>}}\mathrm{<span\; class="notranslate">\; Z</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 0</span>}\\ {\mathrm{<span\; class="notranslate">\; J</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; +</span>\frac{{\mathrm{<span\; class="notranslate">\; l</span>}}_{\mathrm{<span\; class="notranslate">\; 5</span>}}\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; l</span>}}_{\mathrm{<span\; class="notranslate">\; 6</span>}}\mathrm{<span\; class="notranslate">\; Y</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; l</span>}}_{\mathrm{<span\; class="notranslate">\; 7</span>}}\mathrm{<span\; class="notranslate">\; Z</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; l</span>}}_{\mathrm{<span\; class="notranslate">\; 8</span>}}}{{\mathrm{<span\; class="notranslate">\; l</span>}}_{\mathrm{<span\; class="notranslate">\; 9</span>}}\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; l</span>}}_{\mathrm{<span\; class="notranslate">\; 10</span>}}\mathrm{<span\; class="notranslate">\; Y</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; l</span>}}_{\mathrm{<span\; class="notranslate">\; 11</span>}}\mathrm{<span\; class="notranslate">\; Z</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 0</span>}\end{array}\right.<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>$

Among them, I′=II ₀ , J′=JJ ₀ , I, J are the image space coordinates, I ₀ , J ₀ are the coordinate values of the principal point in the image space, l ₁ to l ₁₁ are direct linear transformation coefficients, they are A function of the camera's inner and outer orientation elements. Since the calculation does not require the initial value of the inner orientation element and the outer orientation element, it has become the most basic formula for photogrammetry using ordinary digital cameras.

Changing formula (1) into correction number equation can be transformed into formula (2):

$\left\{\begin{array}{c}<span\; class="notranslate">\; -</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; A</span>}}<span\; class="notranslate">\; [</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; l</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; l</span>}}_{\mathrm{<span\; class="notranslate">\; 9</span>}}{\mathrm{<span\; class="notranslate">\; I</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; )</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; +</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; l</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; l</span>}}_{\mathrm{<span\; class="notranslate">\; 10</span>}}{\mathrm{<span\; class="notranslate">\; I</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; )</span>\mathrm{<span\; class="notranslate">\; Y</span>}<span\; class="notranslate">\; +</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; l</span>}}_{\mathrm{<span\; class="notranslate">\; 3</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; l</span>}}_{\mathrm{<span\; class="notranslate">\; 11</span>}}{\mathrm{<span\; class="notranslate">\; I</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; )</span>\mathrm{<span\; class="notranslate">\; Z</span>}<span\; class="notranslate">\; +</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; l</span>}}_{\mathrm{<span\; class="notranslate">\; 4</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; I</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span><span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; v</span>}}_{{\mathrm{<span\; class="notranslate">\; I</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}\\ <span\; class="notranslate">\; -</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; A</span>}}<span\; class="notranslate">\; [</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; l</span>}}_{\mathrm{<span\; class="notranslate">\; 5</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; l</span>}}_{\mathrm{<span\; class="notranslate">\; 9</span>}}{\mathrm{<span\; class="notranslate">\; J</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; )</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; +</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; l</span>}}_{\mathrm{<span\; class="notranslate">\; 6</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; l</span>}}_{\mathrm{<span\; class="notranslate">\; 10</span>}}{\mathrm{<span\; class="notranslate">\; J</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; )</span>\mathrm{<span\; class="notranslate">\; Y</span>}<span\; class="notranslate">\; +</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; l</span>}}_{\mathrm{<span\; class="notranslate">\; 7</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; l</span>}}_{\mathrm{<span\; class="notranslate">\; 11</span>}}{\mathrm{<span\; class="notranslate">\; J</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; )</span>\mathrm{<span\; class="notranslate">\; Z</span>}<span\; class="notranslate">\; +</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; l</span>}}_{\mathrm{<span\; class="notranslate">\; 8</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; J</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span><span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; v</span>}}_{{\mathrm{<span\; class="notranslate">\; J</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}\end{array}\right.<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>$

Among them: A=l ₉ X+l ₁₀ Y+l ₁₁ Z+1 can iteratively solve formula (2) by the least square method, in order to facilitate fast convergence, take A=1, calculate the 11 coefficients l _i of the photo Approximate value l'_i; Substitute l' _i into the calculation of A value, and then return to (1), iterate repeatedly until the value difference of A is less than the given tolerance, and calculate 11 coefficients l _i and the object space coordinates of the point.

Geometrically calculate the coordinate values of the final actual space, and then connect the corresponding points and lines to establish the three-dimensional numerical model of the outer contour of the reference vehicle and the three-dimensional model of the deformed part of the vehicle involved in the accident, as shown in Figure 3a and Figure 3b , Figure 4a, Figure 4b. Then select 8 points on the front bumper of the main deformation energy-absorbing parts as the characteristic points of the deformation contour, and then use the vector method to calculate and measure the deformation at these 8 characteristic points based on the established benchmark vehicle numerical model , as shown in Figure 5a and Figure 5b, and the results are shown in Table 1. At the same time, it can be measured that the inclination angle between the contour represented by the 8 feature points and the undeformed contour is about 33°.

Table 1 Deformation measurement results of feature points

On the basis of the three-dimensional numerical model obtained by photogrammetry, the collision environment and local finite element model of the deformed vehicle were established, and then the accident site survey information was used to judge that the speed of vehicle A was above 94km/h, and the speed of vehicle B in the accident was not higher than 110km/h, and according to The inclination angles of the deformed contours of the two vehicles obtained by photogrammetry are 33° and 21° respectively, based on which the initial conditions are set, and then iterative calculations are performed.

Through multiple iterative calculations, the deformation values of the two vehicles are obtained when the vehicle A and the vehicle B are traveling in the opposite direction at a direction of 167°, and the speed of the vehicle A is 99.7km/h and the speed of the vehicle B is 101.5km/h at the moment of collision contact. The simulation results are the closest to the real accident results, and the deformation coincidence degree is M=0.9862.

The invention realizes the process reproduction of traffic accidents with no brake marks but vehicle body deformation. Compared with the automobile collision accident reproduction method based on the three-dimensional deformation of key points of the vehicle body, due to the adoption of the photogrammetry method and local finite element technology, the entire accident The reproduction efficiency of the method has been increased by nearly 50%, and the coincidence degree of accident reproduction has reached more than 95% of the original method, which meets the requirements of traffic accident handling.

Claims (8)

1. automobile collision accident reproducting method based on the distortion of photogrammetric and vehicle body outline is characterized in that:

The first step relates to crushed vehicle and and entire car identical with its model carries out photogrammetric to accident;

Described photogrammetric, be meant: evenly put four identical photogrammetric demarcation things at vehicle periphery, form a rectangular area, with this rectangular area is photo centre, with the digital camera clockwise direction vehicle scene is carried out at 4 and find a view, be similar at an angle of 90 between the adjacent shooting orientation, and guarantee to comprise at least in every photo three demarcation things, and four relative positions of demarcating between the thing of difference survey record, realize photogrammetric demarcation to the vehicle scene;

Second step, set up accident vehicle deformation place outline and three-dimensional numerical value model identical with its model and the entire car outline, the photo of the entire car that photographs is imported computing machine, choose the unique point on the appearance profile of vehicle, obtain its pixel coordinate value, set up the relation of image space coordinate system and object space coordinate system according to the DLT method, carry out the object space coordinate that iteration is tried to achieve these points by least square method, then reference point, line are connected, set up the outline model, and as benchmark model;

The 3rd step, analyze benchmark model and distorted pattern, obtain unique point place deflection size and angle of inclination with vector method;

The 4th step, set up the finite element model at involved vehicle deformation position in scene of the accident environment and the accident, concrete steps comprise: to head-on crash or angular impact, the body forward structure contour feature is wanted refinement, to side collision, the side parts of vehicle body will be described in detail;

The 5th step by repeatedly iteration optimization calculating, obtained under setting speed and collision angle condition, and the numerical simulation result of unique point distortion is consistent with true collision measurement result, thereby accurately determined accident generation car speed and collision angle constantly.

2. according to claim 1 based on automobile collision accident the reproducting method photogrammetric and distortion of vehicle body outline, it is characterized in that, the described three-dimensional numerical value model of setting up accident vehicle deformation place outline, be meant: the photo of the crushed vehicle that photographs is imported computing machine, choose point on the deformable contour of vehicle as unique point, obtain its pixel coordinate value, set up the relation of image space coordinate system and object space coordinate system according to the DLT method, carry out the object space coordinate that iteration is tried to achieve these points by least square method, then reference point, line connects, set up the three-dimensional numerical value model of vehicle deformation position outline, these unique points are described the deformable contour of vehicle, are chosen near the automobile frame.

3. according to claim 1 based on automobile collision accident the reproducting method photogrammetric and distortion of vehicle body outline, it is characterized in that, described analysis benchmark model and distorted pattern, obtain the angle of inclination of unique point place deflection size and deformable contour with vector method, be meant: the feature of not distortion place that exists on the vehicle body by distortion and the not crushed vehicle characteristic of correspondence of same model are mated, and then the base value value model is imported in the model of crushed vehicle, the relative position of crushed element in intact car load determined like this, pick up the unique point of deformation place and the corresponding point coordinate in the intact model then, calculate the deflection at each unique point place, and then calculate the angle of inclination of deformable contour.

4. according to claim 1 or 3 described automobile collision accident reproducting methods based on the distortion of photogrammetric and vehicle body outline, it is characterized in that, the described angle of inclination of obtaining unique point place deflection size and deformable contour with vector method, be specially: in the model after coupling, choose two unique point a, b that can represent deformation tendency, suppose that its coordinate is respectively (x _a, y _a, z _a), (x _b, y _b, z _b), some a ', b ' coordinate before the corresponding not distortion in benchmark model are respectively

Think that then unique point a, the b deflection in accident wherein is respectively:

Then the angle of inclination of deformable contour is:

5. according to claim 1 based on automobile collision accident the reproducting method photogrammetric and distortion of vehicle body outline, it is characterized in that, described the 4th step, be meant: by vehicle body design drawing or three-dimensional numerical value model to set up in second step, set up the corresponding finite element model in vehicle deformation position, refinement produces the grid of deformation place, and element length is less than or equal to 10mm, directly replaces with rigid body from distortion zone far away.

6. according to claim 1 based on automobile collision accident the reproducting method photogrammetric and distortion of vehicle body outline, it is characterized in that, described the 5th step, be meant: at first reconnoitre the gained evidence and carry out analysis-by-synthesis according to the scene of the accident, the preliminary scope of determining car speed and collision angle, in this scope with speed and angle as independent variable, carry out iterative computation according to setting step-length, each iterative computation obtains the deflection at character pair point place in one group of finite element model, then according to distortion more photogrammetric result of goodness of fit formula and numerical simulation result, when the deflection of unique point in numerical simulation result and the true casualty effect near the time, think that then pairing impact velocity of this analog result and collision angle are real impact velocity and collision angle.

7. the automobile collision accident reproducting method based on the distortion of photogrammetric and vehicle body outline according to claim 6 is characterized in that, described distortion goodness of fit M formula is specific as follows:

M＝1-(∑(x _iP-x _iA) ²/(∑x _iP·x _iA))

Wherein, x _IPBe the practical distortion amount of the unique point of photogrammetric gained, x _IABe the deflection calculated value that characteristic point position obtains by numerical simulation, the M value is big more, illustrates that the accident reconstruction result is more near truth.

8. according to claim 7 based on automobile collision accident the reproducting method photogrammetric and distortion of vehicle body outline, it is characterized in that, described distortion goodness of fit M value, when numerical simulation result when the legitimate reading, the M value levels off to " 1 ", and the existing constant interval of M value is [m, 1], wherein m be on the occasion of or negative value, concrete numerical value is by the degree of closeness decision of simulation result with true collision accident.

Images