Active and passive security fusion evaluation weight coefficient determination method and system

Abstract

The invention discloses an active and passive safety fusion evaluation weight coefficient determination method and system, and relates to the technical field of automobile safety evaluation, and the method comprises the steps: obtaining active safety data and passive safety data; combining the active and passive data to obtain total sample data; constructing an analytic hierarchy model and a judgment matrix of each hierarchy; calculating a total level sequence of all indexes in the scheme layer to obtain a weight value of each index; performing weighted scoring on each data sample to obtain a total score; constructing a multi-layer perceptron model; and extracting the final weight of each index. According to the method, the vehicle speed after braking of the vehicle active safety system is taken as the collision speed of human-vehicle collision simulation in the passive safety system, and an active safety index and a passive safety index are associated to carry out active and passive safety fusion evaluation; and the index weight coefficient value for fusion evaluation is extracted through an analytic hierarchy process and construction of a multi-layer perceptron model, so that the data accuracy is high, the objectivity is high, and the comprehensiveness is high.

Description

Technical Field

[0001] The present application relates to the technical field of automobile safety assessment, and in particular to a method and system for determining active and passive safety fusion assessment weight coefficients. Background Art

[0002] In recent decades, with the increasing density of urban populations in various countries, the gradual improvement of automobile manufacturing technology, and the continuous increase in the number of cars, the safety of people in road traffic has become more and more prominent. The safety of automobiles has received great attention from the automobile industry and governments around the world. With the implementation of road traffic safety regulations, the advancement of ADAS technology, and the innovation of active and passive safety technologies of automobiles, automobile safety technology is developing in the direction of active and passive safety integration, such as the development of AEB systems. For example, automobile safety technology for pedestrian protection has gradually developed from simple AEB intervention to active and passive safety integrated systems that integrate AEB and active shields.

[0003] The problems with the current automobile safety technology evaluation methods are: the automobile safety evaluation standards used internationally (Euro-NCAP, C-NCAP, IIHS, JNCAP, etc.) only have unilateral evaluation methods for active safety or passive safety. At present, there is still a blank in the field of safety performance evaluation of integrated systems that integrate active and passive safety technologies of vehicles. However, with the maturity of autonomous driving technology, active and passive safety integrated systems will gradually become popular. The safety performance of vehicles equipped with active and passive safety technology integration products requires a complete and scientific evaluation method to evaluate; in the existing safety test procedures, the passive safety evaluation and the active safety evaluation are independent of each other, and the evaluation procedure calculation is a simple superposition of active and passive safety scores, while intelligent safety does not involve the organic integration of active and passive safety, that is, there is no evaluation method that considers the joint evaluation of active and passive safety, and it is impossible to re-determine the weight of each test indicator and the respective weights of active safety and passive safety, resulting in the inability to calculate the total score of the active and passive joint evaluation.

[0004] Therefore, in order to solve the above problems, it is necessary for us to design a reasonable method and system for determining the weight coefficient of active and passive safety fusion evaluation. Summary of the invention

[0005] The present invention provides a method and system for determining weight coefficients of active and passive safety fusion evaluation, which uses the vehicle speed after braking by the vehicle active safety system as the collision speed of human-vehicle collision simulation in the passive safety system, links the active safety index with the passive safety index, and conducts active and passive safety fusion evaluation. The weight coefficient values ​​of the indexes used for fusion evaluation are extracted through the hierarchical analysis method and the construction of a multi-layer perceptron model, and the data has high accuracy, strong objectivity and strong comprehensiveness.

[0006] In order to achieve the above object, the present invention adopts the following technical solutions: In a first aspect, an embodiment of the present application provides a method for determining an active and passive safety fusion evaluation weight coefficient, the method comprising: S100: Obtaining braking data of the vehicle's active safety system in different scenarios and damage data of human-vehicle collision in the passive safety system; Among them, the vehicle speed after braking by the vehicle active safety system is used as the collision speed of the human-vehicle collision simulation in the passive safety system, and the damage data of the human-vehicle collision under different working conditions in the human-vehicle collision simulation environment are obtained, and the damage data of the human-vehicle collision in the passive safety system is obtained; S200: combining the braking data of the active safety system and the damage data of the collision between the person and the vehicle in the passive safety system to obtain total sample data; S300: Constructing a hierarchical analysis model and a judgment matrix of each level in the hierarchical analysis model; S400: Calculate the total hierarchical ranking of all indicators in the solution layer to obtain the weight value of each indicator in the hierarchical analysis model; S500: According to the weight value of each indicator in the hierarchical analysis model, a weighted score is given to each data sample in the total sample data in step S200 to obtain a total score; S600: Based on the total sample data and the weight values ​​of each indicator in the hierarchical analysis model, a multi-layer perceptron model corresponding to the number of indicators is constructed; S700: Extract the output layer weights of the multi-layer perceptron model as the comprehensive weights of the active and passive fusion evaluation.

[0007] As a preferred embodiment of the present invention, in step S200, the active safety system braking data includes at least two variables of deceleration and vehicle speed after braking, wherein the number of variables of deceleration is q1; the number of variables of vehicle speed after braking is q2; The damage data of human-vehicle collision in the passive safety system includes at least five variables of collision position, human body model, vehicle model, engine hood status, and vehicle body pitch angle. The number of variables of collision position is p1; the number of variables of human body model is p2; the number of variables of vehicle model is p3; the number of variables of engine hood status is p4; the number of variables of vehicle body pitch angle is p5; The total number of sample data is at least p1*p2*p3*p4*p5*q1*q2.

[0008] As a preferred embodiment of the present invention, when executing step S300, in the hierarchical analysis model, the target layer is a comprehensive score of active and passive fusion evaluation; the criterion layer uses whether a collision occurs between a vehicle and a pedestrian as the criterion for the next layer of solutions; and the solution layer includes m passive safety indicators and n active safety indicators.

[0009] As a preferred embodiment of the present invention, after executing step S300, a consistency check is performed on the judgment matrix. If the judgment matrix consistency check passes, step S400 is executed; otherwise, an alarm is issued.

[0010] As a preferred embodiment of the present invention, when executing step S600, the multi-layer perceptron model includes at least three layers of nodes: an input layer, an output layer and at least one hidden layer; The input layer contains each data sample in the total sample data; the hidden layer contains the weight value of each indicator in the hierarchical analysis model; and the output layer contains the total score of each data sample.

[0011] As a preferred embodiment of the present invention, after executing step S600, the hyperparameters of the multilayer perceptron model are subjected to Bayesian optimization, specifically: S601: Initialization, select the initial hyperparameter set θ1,…,θ n , and evaluate the objective function f(θ i ); S602: Gaussian process model, using the initial data set D = {(θ i ,f(θ i ))} to train the Gaussian process model; S603: Select an acquisition function and define an acquisition function α(f(θ)); S604: Iterative optimization, in each iteration, select the hyperparameter θ that maximizes the acquisition function new , and evaluate the objective function f(θ new ); S605: Update the model, add the new hyperparameters and objective function values ​​to the data set D, and update the Gaussian process model; S606: Termination condition, repeating steps S603 to S605 until the termination condition is met, that is, the maximum number of iterations or the target performance is reached; S607: Select the best hyperparameter, and select the hyperparameter with the highest posterior probability or objective function value from all evaluated hyperparameters as the final selection.

[0012] As a preferred embodiment of the present invention, when executing step S700, specifically: 80% of the quantitative data samples of the total sample data in step S200 are randomly selected as a training set, and the remaining 20% ​​of the quantitative data samples are used as a verification set, and the hyperparameters of the multilayer perceptron model obtained in step S607 are used to input data to train and verify the model, and the output layer weights of the multilayer perceptron model are extracted as the comprehensive weights of the active and passive fusion evaluation.

[0013] In a second aspect, an embodiment of the present application provides a system for determining active and passive safety fusion evaluation weight coefficients, the system comprising: Data acquisition module; Data combination module; First model building module; A first computing module; Rating calculation module; The second model building module; Final weight extraction module; The data acquisition module obtains the braking data of the vehicle's active safety system under different scenarios and the damage data of the human-vehicle collision in the passive safety system; wherein, the vehicle speed after the vehicle's active safety system is braked is used as the collision speed of the human-vehicle collision simulation in the passive safety system, and the damage data under different working conditions in the human-vehicle collision simulation environment are obtained, and the damage data of the human-vehicle collision in the passive safety system is obtained; the data combination module combines the braking data of the active safety system and the damage data of the human-vehicle collision in the passive safety system to obtain the total sample data; the first model construction module constructs a hierarchical analysis model and the judgment matrix of each level in the hierarchical analysis model; the first calculation module calculates the total hierarchical ranking of all indicators in the scheme layer and obtains the weight value of each indicator in the hierarchical analysis model; the score calculation module performs weighted scoring on each data sample in the total sample data according to the weight value of each indicator in the hierarchical analysis model and obtains the total score; the second model construction module constructs a multi-layer perceptron model corresponding to the number of indicators based on the total sample data and the weight value of each indicator in the hierarchical analysis model; the final weight extraction module trains and optimizes the multi-layer perceptron model, and extracts the output layer weight of the multi-layer perceptron model as the comprehensive weight of the active and passive fusion evaluation.

[0014] In a third aspect, an embodiment of the present application provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the method provided in any possible implementation manner of the first aspect is implemented.

[0015] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the method provided in any possible implementation manner of the first aspect is implemented.

[0016] The beneficial effects of the present invention are as follows: the present invention provides a method and system for determining weight coefficients of active and passive safety fusion evaluation, taking the vehicle speed after braking by the vehicle active safety system as the collision speed of the human-vehicle collision simulation in the passive safety system, linking the active safety index with the passive safety index, and performing active and passive safety fusion evaluation, and extracting the index weight coefficient values ​​for fusion evaluation through the hierarchical analysis method and the construction of a multi-layer perceptron model, with high data accuracy, strong objectivity and strong comprehensiveness. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0018] Figure 1 A flowchart of a method for determining active and passive safety fusion evaluation weight coefficients provided in an embodiment of the present application; Figure 2 A schematic diagram of the structure of a hierarchical analysis model in a method for determining active and passive safety fusion evaluation weight coefficients provided in an embodiment of the present application; Figure 3 A schematic diagram of the structure of a multi-layer perceptron model in a method for determining active and passive safety fusion evaluation weight coefficients provided in an embodiment of the present application. DETAILED DESCRIPTION

[0019] The technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application.

[0020] In the following introduction, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance. The following introduction provides multiple embodiments of the present application, and different embodiments can be replaced or combined, so the present application can also be considered to include all possible combinations of the same and / or different embodiments recorded. Therefore, if one embodiment includes features A, B, and C, and another embodiment includes features B and D, then the present application should also be considered to include embodiments containing one or more of all other possible combinations of A, B, C, and D, although the embodiment may not be clearly recorded in the following text.

[0021] The following description provides examples and does not limit the scope, applicability or examples set forth in the claims. Changes may be made to the functions and arrangements of the elements described without departing from the scope of the present application. Various processes or components may be appropriately omitted, substituted or added to each example. For example, the described method may be performed in an order different from the order described, and various steps may be added, omitted or combined. In addition, features described in some examples may be combined in other examples.

[0022] Example 1: See Figures 1 to 3 , Figure 1 : is a flow chart of a method for determining active and passive safety fusion evaluation weight coefficients provided in an embodiment of the present application. In an embodiment of the present application, the method includes: S100: Obtaining braking data of the vehicle's active safety system in different scenarios and damage data of human-vehicle collision in the passive safety system; Here, the vehicle active safety system refers to the automatic braking system (AEB, Autonomous Emergency Braking). AEB is mainly composed of three modules, including the control module (ECU), the distance measurement module, and the braking module. The core of the distance measurement module includes microwave radar, face recognition technology and video system, etc. It can provide safe, accurate and real-time images and road conditions information on the road ahead. The AEB system uses radar to measure the distance to the vehicle in front or obstacles, and then uses the data analysis module to compare the measured distance with the alarm distance and the safety distance. When it is less than the alarm distance, an alarm will be issued. When it is less than the safety distance, even if the driver does not have time to step on the brake pedal, the AEB system will start to intervene and automatically brake the car, thereby ensuring safe travel.

[0023] In addition, the vehicle's passive safety system focuses on testing the vehicle's protection performance for pedestrians in the event of a collision. That is, it simulates different collision contact shapes, angles and forces when the vehicle collides with pedestrians, and obtains injury data on different parts of the pedestrian's body, so as to provide optimal protection for pedestrians in the event of a vehicle collision.

[0024] Different scenarios include different initial speeds of the vehicle and different road conditions. The active safety system braking data after AEB intervention during intelligent collision avoidance in a certain scenario is obtained, as well as the damage data of the simulated vehicle model and the simulated human model under different working conditions in the scenario. Different working conditions include different vehicle parameters and different human parameters.

[0025] In order to integrate the evaluation of active safety systems and passive safety systems, a common indicator is needed as a medium. The active safety system and the passive safety system have the same indicator, namely the "vehicle speed after braking" of the active safety system and the "pedestrian-vehicle collision speed" of the passive safety system, that is, when AEB intervenes, the vehicle collides with the pedestrian at this speed (or the vehicle stops in front of the pedestrian, at which time the speed is zero).

[0026] Therefore, in the present invention, the vehicle speed after braking by the vehicle active safety system is used as the collision speed of the human-vehicle collision simulation in the passive safety system, the active safety system and the passive safety system are integrated and evaluated, and the damage data under different working conditions in the human-vehicle collision simulation environment are obtained, and the damage data of the human-vehicle collision in the passive safety system is obtained, and the damage data includes vehicle body damage data and human body injury data.

[0027] S200: combining the braking data of the active safety system and the damage data of the collision between the person and the vehicle in the passive safety system to obtain total sample data; In step S200, the active safety system braking data includes at least two variables of deceleration and vehicle speed after braking, where the number of variables of deceleration is q1; the number of variables of vehicle speed after braking is q2; The damage data of human-vehicle collision in the passive safety system includes at least five variables of collision position, human body model, vehicle model, engine hood status, and vehicle body pitch angle. The number of variables of collision position is p1; the number of variables of human body model is p2; the number of variables of vehicle model is p3; the number of variables of engine hood status is p4; the number of variables of vehicle body pitch angle is p5; The total number of sample data is at least p1*p2*p3*p4*p5*q1*q2.

[0028] Here, by effectively combining the data of active safety and passive safety, we can get SQ= p1*p2*p3*p4*p5*q1*q2 data samples for subsequent model calculations. In fact, with the development of technology, the variable dimensions of active safety and passive safety may increase. The active safety system braking data has t-dimensional variables, and the number of variables is q1 to q t ; The passive safety system braking data has y-dimensional variables, and the number of variables is p1 to p y ; Then the total number of sample data SQ is at least p1*p2*…*p y *q1*q2*…*q t .

[0029] Steps S100 and S200 are for obtaining data samples.

[0030] S300: Constructing a hierarchical analysis model and a judgment matrix of each level in the hierarchical analysis model; The Analytic Hierarchy Process (AHP) is a classic subjective weight evaluation method. AHP decomposes the elements related to the evaluation target through the target layer, the criterion layer and the scheme layer, and analyzes them based on the relationships within and between the layers. In this invention, the analytic hierarchy process is used to determine the subjective weight of each test indicator as the initial weight.

[0031] When executing step S300, the hierarchical analysis model is constructed as follows Figure 2 ,In the hierarchical analysis model, the target layer is the comprehensive score of the active and passive ,fusion evaluation; the criterion layer uses whether the vehicle and the pedestrian collide as the criterion of the next layer of ,scheme; the scheme layer includes m passive safety indicators and n active safety indicators.

[0032] Here, among the existing indicators of vehicle collision safety evaluation data, passive safety indicators include impact speed, head injury value (HIC 15 ), maximum bending moment of the thigh (Femur-MAX), maximum bending moment of the calf (Tibia-MAX) and elongation of the knee ligament MCL (ΔMCL); active safety indicators include deceleration (ΔV AEB ), vehicle speed V after AEB system braking af-AEB In the present invention, the impact speed is equal to the vehicle speed V after braking af-AEB .

[0033] Then the solution layer includes Femur-MAX, Tibia-MAX, ΔMCL, HIC15 and V af-AEB These five passive safety indicators and ΔV AEB and V af-AEB These two active safety indicators, but because they have a common V af-AEB Indicators, then the total number of indicators is 5+2-1=6.

[0034] In the judgment matrix of each layer, The judgment matrix of the criterion layer is: ; Among them, CO (Crash Occurred) represents the criterion "collision occurred", NCO (No Crash Occurred) represents the criterion "no collision occurred", and x1 is the importance scale of the CO criterion relative to the NCO criterion. Importance scale: The value is between 1 and 9. If the value is 1, it means that the CO indicator is equally important compared with the NCO indicator; if the value is 9, it means that the CO indicator is extremely important compared with the NCO indicator; The judgment matrix at the solution level is: ; where x ij (i=1, 2, ···, 5; j=1, 2, ···, 5) represents the importance scale of each indicator.

[0035] In the present invention, after executing step S300, a consistency check is performed on the judgment matrix. If the judgment matrix passes the consistency check, step S400 is executed; otherwise, an alarm is issued.

[0036] Here, the consistency index CI (Consistency Index) of each judgment matrix is ​​calculated, and the formula is: ; where λ max is the maximum value of the estimated eigenvalue of the matrix; n is the number of solutions, that is, the order of the matrix; The correspondence table between the number of schemes n and the random consistency parameter RI (Random Index) is as follows: Table 1. Average random consistency parameter (RI) ; Calculate the consistency ratio CR (Consistency Ratio): CR = CI / RI; CI is the consistency index and RI is the consistency parameter. When the CR of the judgment matrix is ​​less than the preset threshold (generally less than 0.10), it is considered that the consistency verification of the judgment matrix has passed, that is, the scale selection of the judgment matrix is ​​reasonable, and there are no major logical errors between the judgments.

[0037] In the present invention, the geometric mean method is used to calculate the composite weights. For a certain judgment matrix, the weight calculation formula is as follows: ; Where W j Represents the AHP weight of the indicators in each judgment matrix. ij are the elements in the matrix (i=1,2,⋯,n).

[0038] S400: Calculate the total hierarchical ranking of all indicators in the solution layer to obtain the weight value of each indicator in the hierarchical analysis model; Here, the AHP weight values ​​of all indicators at the solution level are calculated as: Table 2. AHP weight calculation results ; where w NCO 、w CO Represents the single-level weights of “no collision” and “collision” in the criterion layer; w NCOi (i=1, 2, ···, 5) represents the single-level weight of each solution under the criterion of “no collision”; w COi (i=1, 2, ···, 5) represents the single-level weight of each solution under the criterion of “collision”; W AHPi (i=1, 2, ···, 5) represents the total weight of each solution after weighted calculation, that is, the final initial weight of the indicator. The calculation formula is: W AHPi =w NCO w NCOi +w CO w COi , i=1, 2, …, 6.

[0039] S500: According to the weight value of each indicator in the hierarchical analysis model, a weighted score is given to each data sample in the total sample data in step S200 to obtain a total score; After the weights of each indicator are obtained according to AHP calculation, all samples are weighted and scored to obtain the total score. Among them, the sub-scores of each indicator are determined according to the existing evaluation procedures. Assume that in a certain sample, the score of Femur-MAX is s1; the score of Tibia-MAX is s2; the score of ΔMCL is s3; the score of HIC is s4; the score of ΔMCL is s5; the score of ΔMCL is s6; the score of ΔMCL is s7; the score of ΔMCL is s8; the score of ΔMCL is s9; the score of ΔMCL is s1 ...2; the score of ΔMCL is s3; the score of ΔMCL is s4; the score of ΔMCL 15 The score is s4; ΔV AEB The rating is s5; V af-AEB The score of the sample is s6. Then the total score of the sample is: S = s1W AHP1 + s2W AHP2 +s3W AHP3 +s4W AHP4 +s5W AHP5 +s6W AHP6 .

[0040] Steps S300 to S500 are to calculate weight data for various indicators of vehicle active and passive safety based on the AHP model.

[0041] S600: Based on the total sample data and the weight values ​​of each indicator in the hierarchical analysis model, a multi-layer perceptron model corresponding to the number of indicators is constructed; Based on the simulation data of each sample data in step S200 and the initial weight value of each indicator obtained by using the hierarchical analysis model in step S500, a multi-layer perceptron model matching the number of indicators is constructed; Multilayer Perceptron (MLP) is a feedforward artificial neural network. It contains at least three layers of nodes: an input layer, an output layer, and at least one hidden layer; the input layer contains each data sample in the total sample data; the hidden layer contains the weight values ​​of each indicator in the hierarchical analysis model; the output layer contains the total score of each data sample; in the multilayer perceptron model, except for the input node, each node is a neuron (or processing unit) with a nonlinear activation function; MLP is trained using a supervised learning technique called back propagation.

[0042] The structure of the multi-layer perceptron model constructed by the present invention is as follows: Figure 3 As shown in the figure, W1 to W6 represent the weights of the output layer corresponding to each evaluation index in the MLP network; S1 to S n Represents the comprehensive score of each data sample; that is, the simulation data of each sample data in step S200 is used as the input layer, and the total score is used as the output layer.

[0043] In addition, after executing step S600, the hyperparameters of the multilayer perceptron model are Bayesian optimized. Bayesian optimization: Bayesian optimization is a hyperparameter adjustment method based on Bayesian statistics. It selects hyperparameters that are most likely to improve performance by constructing a probability model of the objective function. This method is particularly suitable for optimization problems with limited computing resources or high evaluation costs. The core idea of ​​Bayesian optimization is to use prior knowledge to guide the search process and update this knowledge in each iteration. In the present invention, the MLP network hyperparameters that need to be optimized are: Learning rate: controls the speed at which the model parameters are updated during training. The number of hidden layers: affects the depth and learning ability of the model. The number of neurons in each layer: affects the complexity and expression ability of the model. Activation function: such as ReLU, Sigmoid or Tanh, affects the nonlinear fitting ability of the model; Bayesian optimization is specifically: S601: Initialization, select the initial hyperparameter set θ1,…,θ n , and evaluate the objective function f(θ i ); S602: Gaussian process model, using the initial data set D = {(θ i ,f(θ i ))} to train the Gaussian process model; The Gaussian kernel function selects the radial basis function (RBF) kernel, which is one of the most commonly used kernel functions. It assumes that the function changes smoothly in the hyperparameter space and decreases rapidly with the increase of distance; the formula is: ; in, is the variance of the signal, is the length scale parameter, which controls the width of the kernel function in the i-th dimension, and d is the dimension of the hyperparameter S603: Select an acquisition function and define the acquisition function α(f(θ)); such as expected improvement (EI) or upper confidence bound (UCB); S604: Iterative optimization, in each iteration, select the hyperparameter θ that maximizes the acquisition function new :θ new = argmax, and evaluate the objective function f(θ new ); S605: Update the model, add the new hyperparameters and objective function values ​​to the data set D, and update the Gaussian process model; S606: Termination condition, repeating steps S603 to S605 until the termination condition is met, that is, the maximum number of iterations or the target performance is reached; S607: Select the best hyperparameter, and select the hyperparameter with the highest posterior probability or objective function value from all evaluated hyperparameters as the final selection.

[0044] S700: Extract the output layer weights of the multi-layer perceptron model as the comprehensive weights of the active and passive fusion evaluation.

[0045] When executing step S700, specifically: randomly extract 80% of the quantitative data samples of the total sample data in step S200 as the training set, and use the remaining 20% ​​of the quantitative data samples as the verification set, according to the hyperparameters of the multilayer perceptron model obtained in step S607, input data to train and verify the model, and extract the output layer weights of the multilayer perceptron model as the comprehensive weights of the active and passive fusion evaluation.

[0046] The present invention discloses a method for determining weight coefficients of active and passive safety fusion evaluation. The method uses the vehicle speed after braking by the vehicle active safety system as the collision speed of human-vehicle collision simulation in the passive safety system, links the active safety index with the passive safety index, conducts active and passive safety fusion evaluation, and extracts the index weight coefficient values ​​used for fusion evaluation by using the hierarchical analysis method and building a multi-layer perceptron model. The method has high data accuracy, strong objectivity and strong comprehensiveness.

[0047] Embodiment 2: A system for determining active and passive safety fusion evaluation weight coefficients provided in the embodiment of the present application includes: Data acquisition module; Data combination module; First model building module; A first computing module; Rating calculation module; The second model building module; Final weight extraction module; The data acquisition module obtains the braking data of the vehicle's active safety system under different scenarios and the damage data of the human-vehicle collision in the passive safety system; wherein, the vehicle speed after the vehicle's active safety system is braked is used as the collision speed of the human-vehicle collision simulation in the passive safety system, and the damage data under different working conditions in the human-vehicle collision simulation environment are obtained, and the damage data of the human-vehicle collision in the passive safety system is obtained; the data combination module combines the braking data of the active safety system and the damage data of the human-vehicle collision in the passive safety system to obtain the total sample data; the first model construction module constructs a hierarchical analysis model and the judgment matrix of each level in the hierarchical analysis model; the first calculation module calculates the total hierarchical ranking of all indicators in the scheme layer and obtains the weight value of each indicator in the hierarchical analysis model; the score calculation module performs weighted scoring on each data sample in the total sample data according to the weight value of each indicator in the hierarchical analysis model and obtains the total score; the second model construction module constructs a multi-layer perceptron model corresponding to the number of indicators based on the total sample data and the weight value of each indicator in the hierarchical analysis model; the final weight extraction module trains and optimizes the multi-layer perceptron model, and extracts the output layer weight of the multi-layer perceptron model as the comprehensive weight of the active and passive fusion evaluation.

[0048] Those skilled in the art can clearly understand that the technical solutions of the embodiments of the present application can be implemented with the help of software and / or hardware. The "unit" and "module" in this specification refer to software and / or hardware that can independently complete or cooperate with other components to complete specific functions, where the hardware can be, for example, a field programmable gate array (FPGA), an integrated circuit (IC), etc.

[0049] It can be understood that the devices provided in the embodiments of the present invention are all applicable to the method described in Embodiment 1. The specific functions of each module can refer to the above method flow and will not be described in detail here.

[0050] Embodiment 3, an electronic device provided by the embodiment of the present invention is used to implement the method described in embodiment 1. The electronic device may include: at least one central processor, at least one network interface, a control interface, a memory, and at least one communication bus.

[0051] Among them, the communication bus is used to realize the connection, communication and information exchange between various components.

[0052] The network interface may optionally include a standard wired interface or a wireless interface (such as a Wi-Fi interface).

[0053] Among them, the control interface is used to output control operations according to instructions.

[0054] The central processing unit may include one or more processing cores. The central processing unit uses various interfaces and lines to connect various parts in the entire terminal, and executes various functions of the terminal and processes data according to the method described in the first embodiment by running or executing instructions, programs, code sets or instruction sets stored in the memory, and calling data stored in the memory.

[0055] The memory may include a random access memory (RAM) or a read-only memory (ROM). Optionally, the memory includes a non-transitory computer-readable storage medium. The memory may be used to store instructions, programs, codes, code sets or instruction sets. The memory may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing an operating system, instructions for at least one function (such as a touch function, a sound playback function, an image playback function, etc.), and methods for implementing the above-mentioned embodiment 1; the data storage area may store data involved in each of the above method embodiments, etc.

[0056] The present invention also provides a computer-readable storage medium, on which a computer program is stored, and when the program is executed by a processor, the method described in the first embodiment is implemented. The computer-readable storage medium may include, but is not limited to, any type of disk, including a floppy disk, an optical disk, a DVD, a CD-ROM, a micro drive, and a magneto-optical disk, a ROM, a RAM, an EPROM, an EEPROM, a DRAM, a VRAM, a flash memory device, a magnetic card or an optical card, a nanosystem (including a molecular memory IC), or any type of medium or device suitable for storing instructions and / or data.

[0057] It should be noted that, for the above-mentioned method embodiments, for the sake of simplicity, they are all described as a series of action combinations, but those skilled in the art should know that the present invention is not limited by the described action sequence, because according to the present invention, certain steps can be performed in other sequences or simultaneously. Secondly, those skilled in the art should also know that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required by the present invention.

[0058] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0059] In the several embodiments provided by the present invention, it should be understood that the disclosed devices can be implemented in other ways. For example, the device embodiments described above are only schematic, such as the division of the units, which is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some service interfaces, and the indirect coupling or communication connection of devices or units can be electrical or other forms.

[0060] The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. In addition, the functional units in the various embodiments of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated units may be implemented in the form of hardware or in the form of software functional units.

[0061] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable memory. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a memory, including several instructions for a computer device (which can be a personal computer, server or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned memory includes: U disk, read-only memory (ROM), random access memory (RAM), mobile hard disk, disk or optical disk, etc., various media that can store program codes.

[0062] A person skilled in the art may understand that all or part of the steps in the various methods of the above embodiments may be completed by entering a program to instruct related hardware, and the program may be stored in a computer-readable memory, which may include a flash drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, etc.

[0063] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the statement "comprise a ..." do not exclude the presence of other identical elements in the process, method, article or device comprising the elements.

[0064] The above contents are further detailed descriptions of the present invention in combination with specific preferred implementation modes, and the specific embodiments of the present invention cannot be considered to be limited to these descriptions. For ordinary technicians in the technical field to which the present invention belongs, several simple deductions or substitutions can be made without departing from the concept of the present invention, which should be regarded as falling within the protection scope of the present invention.

Claims

1. A method for determining active and passive safety fusion evaluation weight coefficients, characterized in that: include: S100: Obtaining braking data of the vehicle's active safety system in different scenarios and damage data of human-vehicle collision in the passive safety system; Among them, the vehicle speed after braking by the vehicle active safety system is used as the collision speed of the human-vehicle collision simulation in the passive safety system, and the damage data of the human-vehicle collision under different working conditions in the human-vehicle collision simulation environment are obtained, and the damage data of the human-vehicle collision in the passive safety system is obtained; S200: combining the braking data of the active safety system and the damage data of the collision between the person and the vehicle in the passive safety system to obtain total sample data; S300: Constructing a hierarchical analysis model and a judgment matrix of each level in the hierarchical analysis model; S400: Calculate the total hierarchical ranking of all indicators in the solution layer to obtain the weight value of each indicator in the hierarchical analysis model; S500: According to the weight value of each indicator in the hierarchical analysis model, a weighted score is given to each data sample in the total sample data to obtain a total score; S600: Based on the total sample data and the weight values ​​of each indicator in the hierarchical analysis model, a multi-layer perceptron model corresponding to the number of indicators is constructed; S700: Extract the output layer weights of the multi-layer perceptron model as the comprehensive weights of the active and passive fusion evaluation.

2. The method for determining active and passive safety fusion evaluation weight coefficient according to claim 1, characterized in that: In step S200, the active safety system braking data includes at least two variables of deceleration and vehicle speed after braking, where the number of variables of deceleration is q1; the number of variables of vehicle speed after braking is q2; The damage data of human-vehicle collision in the passive safety system includes at least five variables: collision position, human body model, vehicle model, engine hood status, and vehicle body pitch angle. The number of variables for the collision position is p1; the number of variables for the human body model is p2; the number of variables for the vehicle model is p3; the number of variables for the vehicle model is p4; the number of variables for the vehicle model is p5; the number of variables for the vehicle model is p6. The number of variables of is p3; the number of variables of the engine hood state is p4; the number of variables of the vehicle body pitch angle is p5; The total number of sample data is at least p1*p2*p3*p4*p5*q1*q2.

3. The method for determining active and passive safety fusion evaluation weight coefficient according to claim 1, characterized in that: When executing step S300, in the hierarchical analysis model, the target layer is the comprehensive score of the active and passive fusion evaluation; the criterion layer uses whether the vehicle and the pedestrian collide as the criterion for the next layer of solutions; the solution layer includes m passive safety indicators and n active safety indicators.

4. The method for determining active and passive safety fusion evaluation weight coefficient according to claim 1, characterized in that: After executing step S300, a consistency check is performed on the judgment matrix. If the judgment matrix passes the consistency check, step S400 is executed; otherwise, an alarm is issued.

5. The method for determining active and passive safety fusion evaluation weight coefficient according to claim 1, characterized in that: When executing step S600, the multilayer perceptron model includes at least three layers of nodes: an input layer, an output layer, and at least one hidden layer; The input layer contains each data sample in the total sample data; the hidden layer contains the weight value of each indicator in the hierarchical analysis model; and the output layer contains the total score of each data sample.

6. The method for determining active and passive safety fusion evaluation weight coefficient according to claim 1, characterized in that: After executing step S600, the hyperparameters of the multilayer perceptron model are optimized by Bayesian optimization, specifically: S601: Initialization, select the initial hyperparameter set θ1,…,θ n , and evaluate the objective function f(θ i ); S602: Gaussian process model, using the initial data set D = {(θ i ,f(θ i ))} to train the Gaussian process model; S603: Select an acquisition function and define an acquisition function α(f(θ)); S604: Iterative optimization, in each iteration, select the hyperparameter θ that maximizes the acquisition function new , and evaluate the objective function f(θ new ); S605: Update the model, add the new hyperparameters and objective function values ​​to the data set D, and update the Gaussian process model; S606: Termination condition, repeating steps S603 to S605 until the termination condition is met, that is, the maximum number of iterations or the target performance is reached; S607: Select the best hyperparameter, and select the hyperparameter with the highest posterior probability or objective function value from all evaluated hyperparameters as the final selection.

7. The method for determining active and passive safety fusion evaluation weight coefficient according to claim 6 is characterized in that: When executing step S700, specifically: randomly extract 80% of the quantitative data samples of the total sample data in step S200 as the training set, and use the remaining 20% ​​of the quantitative data samples as the verification set, according to the hyperparameters of the multilayer perceptron model obtained in step S607, input data to train and verify the model, and extract the output layer weights of the multilayer perceptron model as the comprehensive weights of the active and passive fusion evaluation.

8. A system for determining active and passive safety fusion evaluation weight coefficients, characterized in that: include: Data acquisition module; Data combination module; The first model building module; A first computing module; Rating calculation module; The second model building module; Final weight extraction module; The data acquisition module obtains the braking data of the vehicle's active safety system under different scenarios and the damage data of the human-vehicle collision in the passive safety system; wherein, the vehicle speed after the vehicle's active safety system is braked is used as the collision speed of the human-vehicle collision simulation in the passive safety system, and the damage data under different working conditions in the human-vehicle collision simulation environment are obtained, and the damage data of the human-vehicle collision in the passive safety system is obtained; the data combination module combines the braking data of the active safety system and the damage data of the human-vehicle collision in the passive safety system to obtain the total sample data; the first model construction module constructs a hierarchical analysis model and the judgment matrix of each level in the hierarchical analysis model; the first calculation module calculates the total hierarchical ranking of all indicators in the scheme layer and obtains the weight value of each indicator in the hierarchical analysis model; the score calculation module performs weighted scoring on each data sample in the total sample data according to the weight value of each indicator in the hierarchical analysis model and obtains the total score; the second model construction module constructs a multi-layer perceptron model corresponding to the number of indicators based on the total sample data and the weight value of each indicator in the hierarchical analysis model; the final weight extraction module trains and optimizes the multi-layer perceptron model, and extracts the output layer weight of the multi-layer perceptron model as the comprehensive weight of the active and passive fusion evaluation.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the method according to any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images