Pedestrian injury prediction method in automobile and pedestrian collision accident

Abstract

The invention discloses a pedestrian injury prediction method in an automobile and pedestrian collision accident, which comprises a database establishment stage, a simulation stage and a data processing stage, specifically, a simulation model of an automobile type and each physical sign human body is established, and division is performed according to the collision speed and the collision angle of the human body; performing simulation analysis to obtain human body damaged part and damage degree data, classifying the data, and establishing a database; and processing the data by using a machine learning algorithm, creating an input box, compiling a code for processing the input data, and obtaining a pedestrian damage prediction model, the method of combining simulation and an actual image is adopted to predict an actual collision result, the damage condition of the wounded can be judged in time, and the pedestrian damage prediction accuracy is improved. According to the method, simulation results of most collision accidents are included, high accuracy is achieved for pedestrian damage prediction in the collision accidents, a guiding effect is achieved for rescue personnel to take rescue measures, the requirement for the actual traffic accident data size is greatly reduced, and development of actual application is better facilitated.

Description

Technical Field

[0001] The present invention belongs to the field of collision safety prediction, and specifically relates to a method for predicting damage results in a collision accident between a car and a pedestrian based on a large amount of simulation data. Background Art

[0002] At present, due to the sharp increase in the number of motor vehicles and their use, traffic accidents occur more frequently. The physical injuries of pedestrians are particularly important in the handling of traffic accidents. However, due to the differences in human body signs and the difficulty in classifying traffic accidents, there are great obstacles to the prediction of pedestrian injuries. Summary of the invention

[0003] The purpose of the present invention is to solve the shortcomings of the prior art and to propose a method for predicting pedestrian injuries in a collision accident between a car and a pedestrian.

[0004] In order to achieve the above-mentioned purpose, the present invention adopts the following technical scheme: a method for predicting pedestrian injuries in a collision accident between a car and a pedestrian, comprising: a database establishment stage, a simulation stage and a data processing stage. The database establishment stage creates a simulation model of a vehicle type and a human body with various physical signs, and then divides the human body according to the collision speed and the angle at which the human body is hit, performs simulation analysis to obtain data on the damaged parts of the human body and the degree of damage, and classifies and organizes the data to establish a database; the data processing stage uses a machine learning algorithm to process the data, creates an input box, and writes code to process the input data to obtain a pedestrian injury prediction model.

[0005] As a further description of the above technical solution: the database establishment stage includes collecting pedestrian data, collecting vehicle data, and collecting accident results.

[0006] As a further description of the above technical solution: the collection of pedestrian data includes collecting pedestrian body data, the collection of vehicle data includes collecting and collating data on vehicle model, vehicle weight, length, width and height, and the collection of accident results includes collecting accident results and summarizing and collating pedestrian damage in traffic accidents.

[0007] As a further description of the above technical solution: the simulation stage is divided into three categories. The first category is to model different vehicle models. In view of the simulation time cost, only the area involved in the collision is modeled; the second category is to establish human standing models of different human percentiles, different body shapes, and different postures, and divide the movements of a gait cycle of the human body into five postures for modeling, thereby reducing the simulation time cost; the third category determines the vehicle speed at the time of collision, and divides the vehicle speed into 5km / h or 10km / h, with a value range of 0-80km / h.

[0008] As a further description of the above technical solution: the different human percentiles include the 5th percentile, the 50th percentile and the 95th percentile.

[0009] As a further description of the above technical solution: based on the data of the simulation stage, human standing models of different vehicle types, different vehicle speeds, different human percentiles, different body shapes, and different postures are arranged and combined to obtain simulation results.

[0010] As a further description of the above technical solution: the data processing stage includes a machine learning and code writing stage, in which the simulation results are processed and analyzed to obtain a corresponding prediction model.

[0011] As a further description of the above technical solution: according to the division of different parts of the human body, combined with the data obtained from real accident results in the data collection stage, the degree of damage to various organs and tissues of the human body is determined, and a database of damage conditions of vehicle speed-human percentile-human body shape under different vehicle models is obtained.

[0012] As a further description of the above technical solution: based on the relevant data obtained in the simulation stage, multivariate linear regression analysis is used to determine the correlation between the data, and a prediction model is established. The input is the data of vehicle type, vehicle speed, human percentile, and human body shape, and the output is the location and degree of pedestrian injury in the collision accident.

[0013] Due to the adoption of the above technical solution, the beneficial effects achieved by the present invention are:

[0014] (1) The present invention proposes a method that combines simulation with actual images to predict the actual collision results, which can timely determine the injuries of the injured and provide guidance for rescue personnel to take rescue measures;

[0015] (2) The present invention includes the simulation results of most collision accidents and has a high accuracy in predicting the injuries of pedestrians in collision accidents;

[0016] (3) The method of predicting damage results using simulation proposed in the present invention greatly reduces the demand for the amount of actual traffic accident data and is more conducive to the development of practical applications. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 It is an implementation flow chart of a method for predicting pedestrian injuries in a collision accident between a car and a pedestrian proposed by the present invention. DETAILED DESCRIPTION

[0018] The technical solutions in the embodiments of the present invention are described clearly and completely below. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0019] Reference Figure 1 The present invention provides the following embodiments to describe the present invention in detail:

[0020] Example 1

[0021] In order to solve the problems existing in the prior art, the present invention proposes a method for predicting damage results in a collision accident between a car and a pedestrian based on a large amount of simulation data.

[0022] The method is based on a database obtained from a large amount of simulation data, and a machine learning algorithm is used for data calling and calculation services. The basic data of the accident scene is input to obtain a general situation of pedestrian injuries.

[0023] During the database establishment phase, simulation models of common vehicle models and human bodies with various physical signs are created. The models are then divided according to collision speed and the angle at which the human body is hit. Simulation analysis is performed to obtain data such as the damaged parts of the human body and the degree of damage. The data is then classified and organized to establish a relevant database.

[0024] During the data processing stage, use a suitable machine learning algorithm to process the data, create an input box, and write relevant code to process the input data to obtain a pedestrian injury prediction model. The specific input values ​​should include vehicle model, collision speed, and basic human body signs, such as height and weight.

[0025] The database establishment phase should include the following steps:

[0026] Step 1: Investigate common vehicle models on the market, such as C-class cars, SUVs, MPVs, etc., and establish vehicle simulation models of common sizes for each model.

[0027] Step 2: Establish 5th percentile, 50th percentile and 95th percentile human body models.

[0028] Step 3, classify by vehicle type, human body model, and vehicle speed, use permutation and combination to simulate and obtain the simulation results of each data combination, and extract data such as the human body injury site and injury degree.

[0029] Step 4: classify and organize the data and establish a database, which should include data such as vehicle model, vehicle speed, human body signs and damage conditions.

[0030] Step 5: Use appropriate machine learning algorithms to process the data, obtain the relationship between the above variables, and build a human injury prediction model. The input values ​​should include vehicle type, collision speed, and basic human body signs, such as height and weight.

[0031] Example 2

[0032] In order to make the purpose, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are described in detail below with reference to the accompanying drawings.

[0033] The present invention proposes a method for predicting damage results in a collision accident between a car and a pedestrian based on a large amount of simulation data, which mainly includes data collection, simulation, machine learning and code writing stages.

[0034] The data collection phase is mainly divided into three parts: collecting pedestrian data, collecting vehicle data, and collecting accident results.

[0035] The first part mainly collects pedestrians' physical data, such as height and weight, BMI, body shape and appearance, etc., through volunteers and questionnaires.

[0036] The second part collects and organizes data on the best-selling models on the market, as well as vehicle weight, length, width and height through field surveys and investigations into the sales situation of various automobile manufacturers.

[0037] The third part is to collect accident results, contact public security organs and hospitals, and summarize the damage to pedestrians in various traffic accidents in recent years to facilitate subsequent verification of simulation results.

[0038] In the simulation stage, simulation is mainly classified into three categories.

[0039] The first category is vehicle models. Different vehicle models are modeled. In view of the simulation time cost, etc., only the areas involved in the collision, such as the front bumper, car hood, front windshield and related parts before the A-pillar, can be modeled.

[0040] The second category is to establish human standing models with different percentiles, different body shapes and different postures. The movements of a gait cycle of the human body can be divided into five postures for modeling, which can reduce the simulation time cost.

[0041] The third part is to determine the vehicle speed at the time of collision, and divide the vehicle speed into 5km / h or 10km / h, with a value range of 0-80km / h. Finally, different vehicle models, different speeds, different percentiles of human bodies and different standing postures are arranged and combined to obtain the simulation results.

[0042] The machine learning and code writing phase mainly processes and analyzes the simulation results and obtains the corresponding prediction model. According to the different parts of the human body, combined with the data obtained from the real accident results in the data collection phase, the degree of damage to each organ and tissue of the human body is determined, and a database of damage conditions of vehicle speed-human percentile-human body shape under different vehicle models is obtained.

[0043] According to the relevant data obtained in the simulation stage, multivariate linear regression analysis is used to determine the correlation between the data, and a prediction model is established. The input is the vehicle model, speed, human percentile, human body shape, etc., and the output is the location and degree of pedestrian injury in the collision accident.

[0044] The present invention uses the above embodiments and methods.

[0045] (1) The present invention proposes a method that combines simulation with actual images to predict the actual collision results, which can timely determine the injuries of the injured and provide guidance for rescue personnel to take rescue measures;

[0046] (2) The present invention includes the simulation results of most collision accidents and has a high accuracy in predicting the injuries of pedestrians in collision accidents;

[0047] (3) The method of predicting damage results using simulation proposed in the present invention greatly reduces the demand for the amount of actual traffic accident data and is more conducive to the development of practical applications.

[0048] Finally, it should be noted that the above is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, it is still possible for those skilled in the art to modify the technical solutions described in the aforementioned embodiments or to make equivalent substitutions for some of the technical features therein. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for predicting pedestrian injuries in a collision between a car and a pedestrian, characterized in that: include: The database establishment stage, the simulation stage and the data processing stage. The database establishment stage creates simulation models of vehicle models and human bodies with various physical signs, and then divides the body according to the collision speed and the angle at which the body is hit. Simulation analysis is performed to obtain data on damaged parts of the body and the degree of damage, and the data is classified and organized to establish a database. The data processing stage uses a machine learning algorithm to process the data, creates an input box, and writes code to process the input data to obtain a pedestrian injury prediction model.

2. The method for predicting pedestrian injuries in a collision between a car and a pedestrian according to claim 1, characterized in that: The database establishment stage includes collecting pedestrian data, vehicle data, and accident results.

3. The method for predicting pedestrian injuries in a collision between a car and a pedestrian according to claim 2, characterized in that: The collection of pedestrian data includes collecting pedestrian body data, the collection of vehicle data includes collecting and collating data on vehicle model, vehicle weight, length, width and height, and the collection of accident results includes collecting accident results and summarizing and collating pedestrian damage in traffic accidents.

4. The method for predicting pedestrian injuries in a collision between a car and a pedestrian according to claim 1, characterized in that: The simulation stages are divided into three categories. The first category is to model different vehicle models. In view of the simulation time cost, only the area involved in the collision is modeled; the second category is to establish human standing models of different human percentiles, different body shapes, and different postures, and divide the movements of a gait cycle of the human body into five postures for modeling, thereby reducing the simulation time cost; the third category determines the vehicle speed at the time of collision, and divides the vehicle speed into 5km / h or 10km / h, with a value range of 0-80km / h.

5. The method for predicting pedestrian injuries in a collision between a car and a pedestrian according to claim 4, characterized in that: The different human percentiles include the 5th percentile, the 50th percentile and the 95th percentile.

6. The method for predicting pedestrian injuries in a collision between a car and a pedestrian according to claim 5, characterized in that: According to the data of the simulation stage, human standing models of different vehicle types, different vehicle speeds, different human percentiles, different body shapes, and different postures are arranged and combined to obtain simulation results.

7. The method for predicting pedestrian injuries in a collision between a car and a pedestrian according to claim 6, characterized in that: The data processing stage includes a machine learning and code writing stage, in which the simulation results are processed and analyzed to obtain a corresponding prediction model.

8. The method for predicting pedestrian injuries in a collision between a car and a pedestrian according to claim 7, characterized in that: The human body is divided into different parts and combined with the data obtained from real accident results during the data collection phase to determine the degree of damage to various organs and tissues in the human body, and a database of damage conditions of vehicle speed-human percentile-human body shape under different vehicle models is obtained.

9. The method for predicting pedestrian injuries in a collision between a car and a pedestrian according to claim 8, characterized in that: Based on the relevant data obtained in the simulation stage, multivariate linear regression analysis is used to determine the correlation between the data, and a prediction model is established. The input is the data of vehicle type, vehicle speed, human percentile, and human body shape, and the output is the location and degree of pedestrian injury in the collision accident.

Images