Intelligent network connection automobile safety control method and system based on active and passive safety information fusion

Abstract

The invention discloses an intelligent network connection automobile safety control method and system based on active and passive safety information fusion. The system comprises an environment detection module, a risk prediction module and a cooperative control module. Wherein the environment detection module comprises a passenger state monitoring part, a vehicle system monitoring part and a traffic environment sensing part, the risk prediction module comprises a vehicle collision risk prediction part and a passenger injury risk prediction part, and the cooperative control module comprises an in-vehicle intelligent restraint system, a vehicle steering system and a control execution part of a braking system. The vehicle collision risk probability and the passenger injury probability are calculated in real time according to the established vehicle risk prediction model and the passenger injury risk prediction model by fully utilizing the early warning capability of an active safety system for potential risks. When the system encounters unavoidable dangerous working conditions, the system adjusts the state of the restraint system to enable the restraint system to reach the optimal state of current passenger protection, and extra risks, caused by non-cooperation of the active safety system and the restraint system, of passengers are avoided.

Description

Technical Field

[0001] The present invention relates to a vehicle safety control method and system based on active and passive safety information fusion of intelligent networked vehicles, belonging to the technical field of vehicle safety control and occupant protection. Background Art

[0002] Vehicle safety has always been one of the issues that researchers in the automotive field pay close attention to. Automotive safety technology can be divided into passive safety technology and active safety technology. Vehicle passive safety systems refer to a series of equipment and technologies that reduce or avoid injuries when an accident occurs. These systems do not involve accident prevention, but protect vehicle occupants and pedestrians when an accident occurs. Active safety technology refers to the technology that uses cameras, radars and other equipment to perceive potential dangers in advance and intelligently control the braking and steering systems to automatically avoid collisions, such as the automatic emergency braking system (AEBS). Both technologies are the safety line of the occupants. However, there is a certain degree of fragmentation in the current active and passive safety technologies. Although active safety systems can avoid collision accidents or reduce the effective speed of vehicles when they collide, different collision avoidance control strategies will produce a large braking deceleration under certain typical traffic conditions, causing the occupants to lean forward and leave their positions. This departure reduces the protective effectiveness of the traditional restraint system and increases the risk of secondary collisions for the occupants in the car. Especially when the occupants' initial sitting position is in an out-of-position state, the large braking deceleration will further aggravate the secondary injury of the occupants in the car.

[0003] Therefore, it is increasingly important to study the active and passive safety integrated system that works in synergy. On the one hand, the vehicle restraint system that integrates the perception information of the active safety system can be more adapted to the occupant status and traffic scenes, and improve the protection efficiency of the occupants; on the other hand, the active safety system can intelligently adjust the time and intensity of action to avoid negative impacts on the restraint system. In addition, due to the current level of technological development and the improper behavior of other traffic participants, it is unrealistic for smart cars to completely avoid collision accidents by relying on active safety technologies such as AEBS systems. In this context, it is particularly important to study the effectiveness of the occupant protection system after the active safety system takes effect. Summary of the invention

[0004] In view of the problems to be solved in the above background, the present invention aims to propose a vehicle safety control method and system based on the active and passive safety information fusion of intelligent networked vehicles, so as to make full use of the early warning capability of the active perception system for risks, and calculate the vehicle collision risk probability and occupant injury probability in real time. When encountering unavoidable dangerous working conditions, the system adjusts the state of the restraint system to achieve the optimal state of current occupant protection, avoiding the risk of additional occupant injury caused by the incoordination of the active safety system and the restraint system.

[0005] In order to achieve the above object, the technical solution adopted by the present invention is as follows:

[0006] The vehicle safety control method and system based on the fusion of active and passive safety information of intelligent networked vehicles takes into account the revolutionary changes in the way passengers ride in the future due to the progress of autonomous driving technology and intelligent cockpit technology, and passengers will have a more comfortable and free riding style. This change also puts higher requirements on the active safety technology and passive safety system of the vehicle. The control system includes an environmental detection module, a risk prediction module and a collaborative control module, which provides better protection for the passengers of the autonomous driving vehicle through reasonable control of the vehicle. Among them, the environmental detection module includes three parts: passenger status monitoring, vehicle system monitoring and traffic environment perception, the risk prediction module includes two parts: vehicle collision risk prediction and passenger injury risk prediction, and the collaborative control module includes the control execution part of the vehicle intelligent restraint system and the active safety system (such as the vehicle automatic emergency steering system and automatic braking system).

[0007] The control method comprises the following steps:

[0008] Step 1: The environment detection module obtains the vehicle's own information through the occupant status monitoring module and the vehicle system monitoring module, including the restraint system status, occupant status information and own driving status information (including position, speed and acceleration information);

[0009] Step 2: The environment detection module obtains information about the vehicle's driving environment through the traffic environment perception module, including the kinematic information and shape characteristics of traffic participants such as vehicles, non-motor vehicles and pedestrians around the driving vehicle, and determines the additional risk factor of the target participant based on the pre-established hazard source classification database;

[0010] Step 3: The risk prediction module determines the vehicle collision risk based on the environmental perception information, calculates the minimum acceleration value required for collision avoidance and the vehicle speed at the target location after adopting this collision avoidance strategy; the occupant injury risk prediction module estimates the occupant injury risk probability based on the current minimum acceleration value required for collision avoidance, the vehicle speed at the target location, the additional hazard information of the target participant, the restraint system status, the occupant status information and the vehicle's own driving status information;

[0011] Step 4: The system coordination control module calculates reasonable control parameters of the vehicle intelligent restraint system and active safety system based on the above information.

[0012] Furthermore, the occupant status monitoring module in step 1 includes a vehicle interior camera and pressure sensors on the seat cushion and the seat back, which record and analyze the occupant posture information and the seat belt usage status information in real time; the vehicle system monitoring module includes an angle sensor at the seat hinge, a seat front and rear displacement sensor, a vehicle speed sensor at the power and transmission system, and an angle sensor in the steering system, which obtains the occupant riding status, vehicle speed, vehicle acceleration and front wheel angle information in real time.

[0013] The occupant posture information acquisition is realized based on multimodal data fusion of image data and pressure sensor data. The occupant status monitoring module uses a neural network model to process the occupant's real-time image data and identify human body posture; the pressure sensor installed on the seat feeds back the contact status information between the occupant and the seat in real time, and the information of both parties is mutually verified to increase the recognition success rate.

[0014] The acquisition of the kinematic information of the vehicle is to detect the vehicle's driving speed in the vehicle driving scene, and estimate its driving trajectory in the future based on the current state information of the vehicle. The characterization parameters include the following categories: Where T self (t) represents the position of the vehicle after time t, v self , a self are the vehicle speed and acceleration information, θ self , The vehicle's current steering wheel angle input and steering wheel angle angular velocity information.

[0015] Furthermore, the traffic environment perception module in step 2 includes an outside camera and a radar, wherein the camera is installed on the front windshield to collect real-time images in front and on both sides during driving, obtain shape characteristics of traffic participants, and identify object types; the radar is installed at the front of the vehicle to detect kinematic information such as the relative distance and relative speed between the vehicle and the obstacles in front.

[0016] The traffic participant information acquisition refers to detecting the movement trajectory, movement speed and obstacle type of obstacles around the vehicle. The characterization parameters include: veh (t) = {v veh ,P veh ,a veh , C veh}, where T veh (t) represents the position of other obstacles after time t, P veh Represents the current position coordinate information of the obstacle, v self , a self are the obstacle speed and acceleration information, C vehRepresents the additional risk factor of obstacles with different characteristics. The additional risk factor of the detected target obstacle is determined by the hazard source classification database. One method for establishing the hazard source classification database is to classify traffic participants or obstacles with different characteristics by main features such as size, shape characteristics and color characteristics, and then train the CNN convolutional neural network, and use the trained CNN neural network model to process the information collected by the external camera to complete the identification and classification of potential collision targets in front, and then determine its additional risk factor according to the collision response characteristics of each type of target.

[0017] Furthermore, the prediction of vehicle collision risk in step 3 is to determine the collision risk based on the communication interval of the intersection of the ego vehicle and the obstacle trajectory, t col is the nearest collision risk time node; the current risk of the vehicle is determined by the minimum acceleration required for vehicle collision avoidance; the parameter R col Characterizes the magnitude of the collision risk faced. This parameter can be obtained from the following equation:

[0018]

[0019] Among them, a x_n 、a y_n are the minimum longitudinal and lateral accelerations required to push the collision avoidance time point beyond the risk time threshold, R col is the vehicle collision risk parameter, F lane is the feasibility judgment of lane change, 0 means that lane change is not allowed, 1 means that lane change is allowed, a x_max , a y_max Represents the maximum longitudinal deceleration and lateral acceleration allowed by the current road, K x_n , K y_n They are the longitudinal deceleration risk factor and the lateral acceleration risk factor. When the estimated collision risk exceeds the risk threshold, the vehicle braking system and steering system need to intervene.

[0020] The occupant injury risk prediction module establishes a member injury risk prediction database and then establishes an occupant injury risk prediction model based on machine learning. The occupant injury risk prediction model adopts a hybrid model (Hybrid MLP-SVM model) based on an improved multi-layer perceptron (MLP) and a support vector machine (SVM) to perform prediction tasks for high-dimensional input problems. The first level of the hybrid model is an MLP model, which is divided into an input layer, a hidden layer, and an output layer. The preset occupant injury influencing factors are used as the input of the model. The hidden layer uses a ReLU activation function to improve the model's ability to handle nonlinear problems. The feature vector input by the output layer is passed to the second-level SVM model, and the extracted features are used as the input of the SVM. The SVM is trained to perform classification or regression tasks, and the radial basis kernel function (RBF) is used to handle nonlinear separable situations, and finally the prediction results of occupant injury are obtained.

[0021] The production process of the occupant injury risk prediction database includes the following steps: first, establish a vehicle collision occupant injury value analysis model under the intervention of the active safety system, including the vehicle cockpit model, the occupant restraint system model and the human finite element model, reconstruct a typical accident scene on the Prescan-Carsim-Simulink joint simulation platform, and use the vehicle motion waveform generated during the collision avoidance process as the boundary condition for solving the occupant restraint system model, and analyze the kinematic response of the vehicle during the active safety system intervention process and the occupant injury evaluation index value. In order to ensure the rationality of the sample distribution in the data set, the parameters such as the vehicle collision state, the active safety system action time and intensity, the occupant seat belt usage, the occupant sitting posture, the seat back angle, and the front and rear displacement of the seat are used as variables. Based on the experimental design method, the occupant injury value analysis model under different parameters is simulated and solved, and the dynamic characteristics of the head, neck, chest and other parts of the dummy model are obtained, and an occupant response data set with a sufficient number of samples is established. Among them, 80% of the sample data are used as training samples, and 20% of the sample data are used as test samples. Since there are many influencing parameters in the test, the results are first subjected to Pearson correlation analysis. The calculation method is as follows. Based on the results, the factors that have a greater impact on occupant injuries are determined to simplify the model structure and improve computational efficiency.

[0022]

[0023] Among them, r XY represents the correlation coefficient, σ X , σ Y represents the standard deviation of X and Y. represents the mean of X and Y, n is the number of samples, in this database analysis, X irepresents the above-mentioned influencing factors, such as the vehicle's motion state at the time of collision, the parameters of the active safety system's action time and intensity, the use of the occupant's seat belt, the occupant's sitting posture, the seat back angle, and the seat's front-to-back displacement. i An evaluation index value representing occupant injury. The present invention adopts two major indicators, namely the comprehensive injury evaluation index WIC and the AIS injury score. The head, neck and chest of the occupant are the key focus areas, and the injury evaluation index value is calculated based on the output parameters of this area during the test. WIC is used for vehicle system control optimization. The goal is to minimize the WIC value by reasonably controlling the parameters of the vehicle's braking system, steering system and restraint system. Its coefficients a, b, and c are the statistical probabilities of damage to this area in the simulation test; while the AIS injury score focuses on the risk of the severity of damage to key areas, and its risk level is determined by the probability of death caused by damage to this area. The specific response value of the test model and the AIS score are linked by an empirical formula.

[0024] Furthermore, the coordinated control module in step 4 selects the best driving control strategy in real time based on the information detected and calculated by each part. When the vehicle is in a typical risk scenario such as a blind spot or a complex intersection, the system coordinated control module selects the best driving control strategy in real time based on the vehicle collision risk index R calculated in real time by the risk prediction module. col and occupant injury risk indicators AIS and WIC to formulate appropriate coordinated control strategies. When the vehicle collision risk index exceeds the set threshold, it is determined whether the vehicle's collision avoidance capability can avoid the accident under the current working conditions. If the accident can be avoided, the system will give priority to controlling the vehicle to automatically avoid the collision; if the collision accident is unavoidable, the system will combine the restraint system information, occupant posture information, vehicle motion information and collision target information to formulate control rules for the vehicle's braking and steering systems, adjust the vehicle restraint system status, and reduce the AIS occupant injury risk to a low level; when the AIS score level cannot be reduced, the optimization goal is to reduce the WIC score.

[0025] Beneficial effects of the present invention:

[0026] 1. The current occupant injury prediction model does not consider the impact of the active safety system on the restraint system. The response results obtained by testing according to the ideal restraint system state and occupant sitting posture have certain errors. This paper fully considers the changing factors of the matching relationship between the restraint system and the occupant, and the proposed prediction model training scheme is more comprehensive.

[0027] 2. The present invention proposes an active and passive safety fusion control strategy for inevitable accident scenarios, which can ensure vehicle safety while improving traffic efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 The schematic diagram of the overall principle of the system is shown in Figure 2.

[0029] Figure 2 Indicates the installation position of the camera and pressure sensor in the occupant posture detection solution;

[0030] Figure 3 Schematic diagram of the database and hybrid prediction algorithm representing the occupant injury prediction model. DETAILED DESCRIPTION

[0031] The present invention will be further described below in conjunction with the accompanying drawings.

[0032] Vehicle safety control method and system based on active and passive safety information fusion of intelligent connected vehicles, such as Figure 1 As shown, the control system includes an environment detection module, a risk prediction module and a collaborative control module, which provides better protection for the occupants of the autonomous driving vehicle through reasonable control of the vehicle. Among them, the environment detection module includes three parts: occupant status monitoring, vehicle system monitoring and traffic environment perception, the risk prediction module includes two parts: vehicle collision risk prediction and occupant injury risk prediction, and the collaborative control module includes the control execution part of the in-vehicle intelligent restraint system, vehicle steering system and braking system.

[0033] In order to achieve the above object, the control method comprises the following steps:

[0034] Step 1: Obtain vehicle information through the occupant status monitoring module and the vehicle system monitoring module, including the restraint system status, occupant status information and vehicle driving status information (including position, speed and acceleration information);

[0035] Step 2: Obtain information about the vehicle's driving environment through the traffic environment perception module, including the kinematic information and shape characteristics of traffic participants such as vehicles, non-motor vehicles and pedestrians around the driving vehicle, and determine the additional risk factor of the target participant based on the pre-established hazard source classification database;

[0036] Step 3: Determine the vehicle collision risk based on the environmental perception information, calculate the minimum acceleration value required for collision avoidance and the vehicle speed at the target location after adopting this collision avoidance strategy; the occupant injury risk prediction module estimates the occupant injury risk probability caused by the current minimum acceleration value required for collision avoidance, the vehicle speed at the target location, the additional hazard information of the target participant, the restraint system status, the occupant status information and the vehicle's own driving status information;

[0037] Step 4: Calculate reasonable control parameters of the vehicle intelligent restraint system and active safety system according to the above method.

[0038] In one embodiment, Figure 2As shown, the occupant status monitoring module in step 1 includes a vehicle interior camera and pressure sensors on the seat cushion and the seat back, which record and analyze the occupant posture information and the seat belt usage status information in real time; the vehicle system monitoring module includes an angle sensor at the seat hinge, a seat front and rear displacement sensor, a vehicle speed sensor at the power and transmission system, and an angle sensor in the steering system, which obtains the occupant's riding status, vehicle speed, vehicle acceleration and front wheel angle information in real time.

[0039] The acquisition of the occupant posture information is achieved based on multimodal data fusion of image data and pressure sensor data. The occupant status monitoring module uses the OpenPose model to process the real-time image data of the occupant. The model is an open source human posture recognition model developed based on convolutional neural networks and supervised learning and using the caffe framework. It can generate human joint features in a plane and draw lines between joints to identify whether the occupant is in a normal sitting posture or an abnormal sitting posture. The occupant status monitoring module identifies the contact state between the occupant and the seat through the pressure sensors symmetrically installed on the seat. For example, the system can determine the tilt state of the occupant based on the information of the left and right pressure sensors. The contact state can be subdivided into occupant back out of position, occupant offset, and occupant diving. The occupant status monitoring module simultaneously receives data from the human posture recognition model and the pressure sensor, and the information from both sides is mutually verified to increase the recognition success rate.

[0040] The acquisition of the kinematic information of the vehicle is to detect the vehicle's driving speed in the vehicle driving scene, and estimate its driving trajectory in the future based on the current state information of the vehicle. The characterization parameters include the following categories: Where T self (t) represents the position of the vehicle after time t, v self , a self are the vehicle speed and acceleration information, θ self , The vehicle's current steering wheel angle input and steering wheel angle angular velocity information.

[0041] In one embodiment, the traffic environment perception module in step 2 includes an outside camera and a radar, wherein the camera is installed on the front windshield to collect real-time images of the front and both sides during driving, obtain the shape characteristics of traffic participants, and identify the object type; the radar is installed at the front of the vehicle to detect kinematic information such as the relative distance and relative speed between the vehicle and the obstacles in front.

[0042] The traffic participant information acquisition refers to detecting the movement trajectory, movement speed and obstacle type of obstacles around the vehicle. The characterization parameters include: veh (t) = {v veh ,P veh ,a veh , Cveh}, where T veh (t) represents the position of the obstacle after time t, P veh Represents the current position coordinate information of the obstacle, v self , a self are the obstacle speed and acceleration information, C veh It represents the additional risk coefficient of obstacles with different characteristics. To determine the additional risk coefficient, we must first establish a hazard source identification database, that is, rely on the CNN convolutional neural network model to complete the target identification and classification of traffic participants with different characteristics. The main identification features include their external dimensions, color characteristics, etc. Considering the difference in the risk of damage caused by the structure of the front target to the rear vehicle, the preset labels of traffic participants in the CNN model mainly include non-motor vehicles, passenger cars, regular tail commercial vehicles, irregular tail commercial vehicles, pedestrians and guardrails and other fixed traffic facilities. Then, in the LS-dyna finite element simulation platform, different types of traffic participants such as non-motor vehicles, passenger cars, pedestrians and other finite element models are established to conduct collision simulation tests between the vehicle and different obstacles. Among them, a finite element dummy model is placed at the seat of the vehicle model, and the size of the injury is measured by the dynamic response value of its key parts in the collision simulation experiment. Then, after a large number of tests, the average response result of the vehicle occupant model is obtained. The response results of the occupant model are such as the synthetic acceleration HIC of the head. 15 、Neck injury standard N ij and chest displacement C disp Finally, the test results when the target obstacle type is a passenger car are used as the benchmark 1, and the test results of other types of traffic participants are normalized to obtain the additional risk coefficients of different traffic participants.

[0043] In one embodiment, the prediction of the vehicle collision risk in step 3 is to determine the collision risk based on the communication interval of the intersection of the vehicle and the obstacle trajectory. The intersection is determined by the following method, in which the pedestrian protection factor is taken into consideration, so the pedestrian-vehicle collision threshold is listed separately from other collision risks, T ped_r (t) represents the location information of the pedestrian after time t:

[0044] t i ={t|dis(T self (t),T ped_r (t)) <range_sr&dis(T self (t),T veh (t)) <range_sv}

[0045] t col =min(t i )

[0046] Among them, range_sv and range_sr are the collision thresholds of vehicle-to-vehicle collision and human-to-vehicle collision, which are related to the vehicle's own body parameters and the size of the surrounding vehicles. i is the time point at which the vehicle may have a collision risk in the future, t col is the nearest collision risk time node.

[0047] The risk of the vehicle is determined by the minimum acceleration required for the vehicle to avoid a collision:

[0048] a x_n =min{a x}

[0049] st{a x |t col >t col_r}

[0050] a y_n =min{a y}

[0051] st{a y |t col >t col_r}

[0052] Among them, t col_r is the collision risk time threshold, which is determined by the vehicle's own characteristics; a x_n 、a y_n They are the minimum longitudinal and lateral accelerations required to push the collision avoidance time point beyond the risk time threshold.

[0053] Using parameter R col Characterizes the magnitude of the collision risk faced. This parameter can be obtained from the following equation:

[0054]

[0055] Among them, R col is the vehicle collision risk parameter, F lane is the feasibility judgment of lane change, 0 means that lane change is not allowed, 1 means that lane change is allowed, a x_max , a y_max Represents the maximum longitudinal deceleration and lateral acceleration allowed by the current road, K x_n , K y_n They are the longitudinal deceleration risk coefficient and the lateral acceleration risk coefficient respectively. The risk coefficient needs to be given under comprehensive consideration of traffic efficiency and driving safety. The temporary values ​​are all 1. When more emphasis is placed on driving safety, the coefficient value can be appropriately increased.

[0056] When the estimated collision risk exceeds the risk threshold, the vehicle's braking system and steering system need to intervene. The maximum value of the risk threshold is 1. Taking into account system errors such as the recognition deviation of the maximum lateral acceleration and the maximum longitudinal deceleration, as well as system safety redundancy, this value will be less than 1.

[0057] In one embodiment, the occupant injury risk prediction relies on deep learning to establish an occupant injury risk prediction model and its database, such as Figure 3 shown.

[0058] Among them, the occupant injury risk prediction model is based on the improved multi-layer perceptron (MLP) and support vector machine (SVM) hybrid model (Hybrid MLP-SVM model) to perform prediction tasks for high-dimensional input problems. The model combines the deep feature extraction capability of MLP and the excellent classification performance of SVM, and is suitable for processing complex data sets in high-dimensional feature space. The first level of the hybrid model is the MLP model, which is divided into an input layer, a hidden layer, and an output layer. The preset occupant injury influencing factors are used as the input of the model. The hidden layer uses the ReLU activation function to improve the model's ability to handle nonlinear problems. The feature vector input by the output layer is passed to the second-level SVM model, and the extracted features are used as the input of the SVM. The SVM is trained to perform classification or regression tasks. At the same time, the radial basis kernel function (RBF) is used to handle nonlinear separable situations, and finally the prediction results of occupant injury are obtained.

[0059] The production process of the occupant injury risk prediction database includes the following steps: first, establish a vehicle collision occupant injury value analysis model under the intervention of the active safety system, including a vehicle cockpit model, an occupant restraint system model and a human finite element model, reconstruct a typical accident scene on a simulation platform, and use the vehicle motion waveform generated during the collision avoidance process as the boundary condition for solving the occupant injury value analysis model, and analyze the kinematic response of the vehicle during the active safety system intervention process and the occupant injury evaluation index value. In order to ensure the rationality of the sample distribution in the data set, the parameters such as the vehicle collision state, the active safety system action time and action intensity, the occupant seat belt usage, the occupant sitting posture, the seat back angle, and the seat front and rear displacement are used as variables. Based on the experimental design method, the occupant injury value analysis model under different parameters is simulated and solved, and the dynamic characteristics of the head, neck, chest and other parts of the dummy model are obtained, and an occupant response data set with a sufficient number of samples is established. Among them, 80% of the sample data are used as training samples, and 20% of the sample data are used as test samples. Since there are many influencing parameters in the test, the results are first subjected to Pearson correlation analysis. The calculation method is as follows. Based on the results, the factors that have a greater impact on occupant injuries are determined to simplify the model structure and improve computational efficiency.

[0060]

[0061] Among them, X i represents the above-mentioned influencing factors, Y i The evaluation index value representing the injury of the occupant. The comprehensive injury evaluation index WIC and AIS injury score are adopted. The head, neck and chest of the occupant are the focus areas. The injury evaluation index value is calculated based on the output parameters of the parts in the test. The calculation standards for the response index of the occupant's head, neck and chest are respectively the head composite acceleration HIC 15 、Neck injury standard N ij and chest displacement C disp .

[0062] In the embodiment of the present application, WIC is used for control optimization of the vehicle system. The goal is to minimize the WIC value by reasonably controlling the parameters of the vehicle braking system, steering system and restraint system. The calculation method of this indicator is as follows, and its coefficients a, b, and c are the statistical probabilities of damage to the part in the simulation test:

[0063] WIC=C veh (aHIC 15 +bN ij +cC disp )

[0064] sta+b+c=1

[0065] The AIS injury score focuses on the risk of the severity of injuries to key parts. The risk level is determined by the probability of death caused by the injury to the part. The specific response value of the test model and the AIS score are linked by an empirical formula. The probability of the occupant's head, neck, and chest at each AIS injury level is calculated as follows:

[0066]

[0067] Among them, P(AIS head ≥i), i = 1, 2, 3, 4, 5, represents the probability of head injury exceeding the i-level standard of the AIS injury assessment standard, P(AIS neck ≥i), i = 2, 3, 4, 5, represents the probability that the neck injury exceeds the i-level standard of the AIS injury assessment standard, P(AIS chest ≥i), i = 2, 3, 4, representing the probability that the chest injury exceeds the i-level standard of the AIS injury assessment standard.

[0068] When the vehicle is in a typical risk scenario, the vehicle collision risk index R is calculated in real time. coland occupant injury risk indicators AIS and WIC. When the vehicle collision risk index exceeds the set threshold, it is determined whether the vehicle's collision avoidance capability can avoid the accident under the current working conditions. If the accident can be avoided, the collaborative control module will prioritize controlling the vehicle to automatically avoid the collision. If a collision accident is unavoidable, the collaborative control module will combine the restraint system information, occupant posture information, vehicle motion information and collision target information to formulate control rules for the vehicle's braking system and steering system, adjust the vehicle's intelligent restraint system status, and reduce the AIS occupant injury risk to a low level. When the AIS score level cannot be lowered, reducing the WIC score is the optimization goal.

[0069] The series of detailed descriptions listed above are only specific descriptions of feasible implementation methods of the present invention. They are not intended to limit the scope of protection of the present invention. All equivalent methods or changes that do not deviate from the technical creation of the present invention should be included in the scope of protection of the present invention.

Claims

1. Intelligent connected vehicle safety control system based on active and passive safety information fusion, characterized by: include: Environmental detection module, risk prediction module and collaborative control module; The environment detection module includes three parts: passenger status monitoring module, vehicle system monitoring module and traffic environment perception module; The risk prediction module includes two parts: a vehicle collision risk prediction module and an occupant injury risk prediction module; The collaborative control module includes three parts: a vehicle intelligent restraint system module, an automatic emergency steering system module in the vehicle active safety system, and an automatic emergency braking system module; The environment detection module obtains vehicle information, including restraint system status, occupant status information and vehicle driving status information, including position, speed and acceleration information, through the occupant status monitoring module and the vehicle system monitoring module. On the other hand, it obtains vehicle driving environment information, including kinematic information and shape characteristics of traffic participants such as vehicles, non-motor vehicles and pedestrians around the driving vehicle, through the traffic environment perception module, and determines the additional risk coefficient of the target participant based on the pre-established hazard source classification database. The risk prediction module, based on the environmental perception information, determines the vehicle collision risk through the vehicle collision risk prediction module, calculates the minimum acceleration value required for collision avoidance and the vehicle speed of the vehicle traveling to the target position after adopting the collision avoidance strategy; on the other hand, estimates the minimum acceleration value required for current collision avoidance, the vehicle speed at the target position, the additional danger information of the target participant, the restraint system status, the occupant status information and the occupant injury risk probability caused by the vehicle's own driving status information through the occupant injury risk prediction module; The system coordination control module calculates reasonable control parameters of the vehicle intelligent restraint system and active safety system.

2. The intelligent connected vehicle safety control system based on active and passive safety information fusion according to claim 1 is characterized in that: The occupant status monitoring module includes a vehicle interior camera and pressure sensors on the seat cushion and backrest to record and analyze occupant posture information and seat belt usage status information in real time; Among them, the acquisition of the occupant posture information is achieved based on multimodal data fusion of image data and pressure sensor data, and the neural network model is used to process the real-time image data of the occupant to identify the human body posture; the pressure sensor installed on the seat provides real-time feedback on the contact status between the occupant and the seat, and the information of both parties is mutually verified to increase the recognition success rate.

3. The intelligent networked vehicle safety control system based on active and passive safety information fusion according to claim 1 is characterized in that: The vehicle system monitoring module includes a rotation angle sensor at the seat hinge, a seat front and rear displacement sensor, a vehicle speed sensor at the power and transmission system, and a rotation angle sensor in the steering system, which can obtain the passenger's riding status, vehicle speed, vehicle acceleration and front wheel angle information in real time; The acquisition of ego vehicle kinematic information is to detect the vehicle's driving speed in the vehicle driving scene, and estimate its driving trajectory in the future based on the current state information of the ego vehicle. The characterization parameters include the following categories: Where T self (t) represents the position of the vehicle after time t, v self , a self are the vehicle speed and acceleration information, θ self , The vehicle's current steering wheel angle input and steering wheel angle angular velocity information.

4. The intelligent networked vehicle safety control system based on active and passive safety information fusion according to claim 1 is characterized in that: The traffic environment perception module includes an external camera and a radar, wherein the camera is installed on the front windshield to collect real-time images of the front and both sides during driving, obtain the shape characteristics of traffic participants, and identify the type of object; the radar is installed at the front of the vehicle to detect the relative distance and relative speed between the vehicle and the obstacles in front; The traffic participant information is obtained by detecting the movement trajectory, movement speed and obstacle type of obstacles around the vehicle. The characterization parameters include: veh (t) = {v veh ,P veh ,a veh , C veh }, where T veh (t) represents the position of the obstacle after time t, P veh Represents the current position information of the obstacle, v self , a self are the obstacle speed and acceleration information, C veh It represents the additional risk factor of obstacles with different characteristics. Traffic participants with different characteristics use a neural network model to process the information collected by the external camera to complete the identification and classification of potential collision targets in front. The classification features include their size and color characteristics.

5. The intelligent connected vehicle safety control system based on active and passive safety information fusion according to claim 1 is characterized in that: The vehicle collision risk prediction module determines the collision risk according to the communication interval of the intersection point between the vehicle and the obstacle trajectory, and uses the following method to determine the intersection point: t i ={t|dis(T self (t),T ped_r (t))<range_sr&dis(T self (t),T veh (t))<range_sv} t col =min(t i ) Among them, range_sr and range_sv are the collision thresholds for vehicle-to-vehicle collision and human-to-vehicle collision, respectively. i is the time point at which the vehicle may have a collision risk in the future, t col is the nearest collision risk time node, T ped_r (t) represents pedestrians among obstacles, which are listed separately for the sake of pedestrian protection; The risk of the vehicle is determined by the minimum acceleration required for the vehicle to avoid a collision: in x_n =min{a x } s.t.{a x |t col >t col_r } in y_n =min{a y } s.t.{a y |t col >t col_r } Among them, t col_r is the collision risk time threshold, which is determined by the vehicle's own characteristics; a x_n 、a y_n are the minimum longitudinal and lateral accelerations required to push the collision avoidance time point beyond the risk time threshold, respectively; Using parameter R col Characterizes the size of the collision risk faced. This parameter can be obtained by the following equation: Among them, R col is the vehicle collision risk parameter, F lane is the feasibility judgment of lane change, 0 means that lane change is not allowed, 1 means that lane change is allowed, a x_max , a y_max Represents the maximum longitudinal deceleration and lateral acceleration allowed by the current road, K x_n , K y_n They are the longitudinal deceleration risk factor and the lateral acceleration risk factor respectively.

6. The intelligent networked vehicle safety control system based on active and passive safety information fusion according to claim 1 is characterized in that: The occupant injury risk prediction module establishes a collision accident simulation database and then establishes an occupant injury risk prediction model based on machine learning. The occupant injury value prediction model adopts a hybrid model based on an improved multi-layer perceptron and a support vector machine to predict high-dimensional input problems. The first level of the hybrid model is an MLP model, which is divided into an input layer, a hidden layer, and an output layer. The preset occupant injury influencing factors are used as the input of the model. The hidden layer uses a ReLU activation function to improve the model's ability to handle nonlinear problems. The feature vector input by the output layer is passed to the second-level SVM model, and the extracted features are used as the input of the SVM. The SVM is trained to perform classification or regression tasks. At the same time, a radial basis kernel function (RBF) is used to handle nonlinear separable situations, and finally the prediction result of occupant injury is obtained.

7. The intelligent networked vehicle safety control system based on active and passive safety information fusion according to claim 6 is characterized in that: The collision accident simulation database for occupant injury risk prediction model training and testing is established by the following process: Firstly, an analysis model of occupant injury value in vehicle collision under active safety system intervention is established, including vehicle cockpit model, occupant restraint system model and human finite element model. A typical accident scene is reconstructed on the simulation platform, and the vehicle motion waveform generated during the collision avoidance process is used as the boundary condition for solving the occupant injury value analysis model. The kinematic response of the vehicle during the active safety system intervention process and the occupant injury evaluation index value are analyzed. In order to ensure the rationality of the sample distribution in the data set, the vehicle motion state at the moment of collision, the action time and intensity of the active safety system, the use of the occupant seat belt, the occupant sitting posture, the seat back angle, and the front and rear displacement of the seat are used as variables to simulate and solve the occupant injury value analysis model under different parameters, obtain the dynamic characteristics of the head, neck and chest of the dummy model, and establish an occupant response data set with a sufficient number of samples. Among them, 80% of the sample data are used as training samples, and 20% of the sample data are used as test samples. The results are subjected to Pearson correlation analysis, and the calculation method is as follows. According to the results, the factors that have a greater impact on occupant injury are determined, the model structure is simplified, and the calculation efficiency is improved. Among them, X i represents the above-mentioned influencing factors, Y i An evaluation index value representing occupant damage.

8. The intelligent networked vehicle safety control system based on active and passive safety information fusion according to claim 7 is characterized in that: The collaborative control module selects the best driving control strategy in real time based on the information detected and calculated by each module; when the vehicle is in a typical risk scenario, the system coordinated control module selects the best driving control strategy in real time based on the vehicle collision risk index R calculated in real time by the risk prediction module. col , formulate appropriate coordinated control strategies in combination with the occupant injury risk indicators AIS and WIC; if the vehicle collision risk indicator exceeds the set threshold, determine whether the vehicle's collision avoidance capability can avoid the accident under the current working conditions. If the accident can be avoided, the system will give priority to controlling the vehicle to automatically avoid the collision; If a collision is unavoidable, the system will combine the restraint system information, occupant posture information, vehicle motion information and collision target information to formulate control rules for the vehicle braking and steering systems, adjust the vehicle restraint system status, and reduce the AIS occupant injury risk to a low level; when the AIS score level cannot be lowered, the optimization goal is to reduce the WIC score.

9. The intelligent networked vehicle safety control system based on active and passive safety information fusion according to claim 8 is characterized in that: The WIC is used for control optimization of vehicle systems. The goal is to minimize the WIC value by reasonably controlling the parameters of the vehicle braking system, steering system and restraint system. The calculation method of this indicator is as follows. The coefficients a, b, and c are the statistical probabilities of damage to the part in the simulation test: WIC=C veh (aHIC 15 +bN ij +cC disp ) sta+b+c=1 The AIS injury score focuses on the risk of the severity of injuries to key parts. The risk level is determined by the probability of death caused by the injury to the part. The specific response value of the test model and the AIS score are linked by an empirical formula. The probability of the occupant's head, neck, and chest at each AIS injury level is calculated as follows:

10. The control method of the intelligent networked vehicle safety control system based on active and passive safety information fusion according to any one of claims 1 to 9, characterized in that: The steps include: Step 1: Use the environment detection module to obtain the vehicle's own information through the occupant status monitoring module and the vehicle system monitoring module, including the restraint system status, occupant status information and own driving status information, including position, speed and acceleration information; Step 2: Use the environment detection module to obtain information about the vehicle's driving environment through the traffic environment perception module, including the kinematic information and shape characteristics of traffic participants such as vehicles, non-motor vehicles and pedestrians around the driving vehicle, and determine the additional risk factor of the target participant based on the pre-established hazard source classification database; Step 3: Use the risk prediction module to determine the vehicle collision risk based on the environmental perception information, calculate the minimum acceleration value required for collision avoidance, and the vehicle speed at which the vehicle travels to the target position after adopting this collision avoidance strategy; The occupant injury risk prediction module estimates the occupant injury risk probability caused by the minimum acceleration value required for current collision avoidance, the vehicle speed at the target position, the additional hazard information of the target participant, the restraint system status, the occupant status information and the own driving status information; Step 4: Use the system coordination control module to calculate reasonable control parameters of the vehicle intelligent restraint system, the vehicle automatic emergency steering system and the vehicle automatic emergency braking system based on the above information. Among them, the occupant status monitoring module in step 1 includes a vehicle internal camera and pressure sensors on the seat cushion and backrest, which record and analyze the occupant posture information and seat belt usage status information in real time; the vehicle system monitoring module includes a rotation angle sensor at the seat hinge, a seat front and rear displacement sensor, a vehicle speed sensor at the power and transmission system, and an angle sensor in the steering system, which obtains the occupant's riding status, vehicle speed, vehicle acceleration and front wheel angle information in real time The occupant posture information acquisition is realized based on multimodal data fusion of image data and pressure sensor data. The occupant status monitoring module uses a neural network model to process the occupant's real-time image data and recognize the human body posture; the pressure sensor installed on the seat feeds back the contact status information between the occupant and the seat in real time, and the information of both parties is mutually verified to increase the recognition success rate; The acquisition of the kinematic information of the ego vehicle is to detect the vehicle's driving speed in the vehicle driving scene, and estimate its driving trajectory in the future based on the current state information of the ego vehicle. The characterization parameters include the following categories: Where T self (t) represents the position of the vehicle after time t, v self , a self are the vehicle speed and acceleration information, θ self , The current steering wheel angle input and steering wheel angle angular velocity information of the vehicle; Step 2: The traffic environment perception module includes an external camera and a radar. The camera is installed on the front windshield to collect real-time images of the front and both sides during driving, obtain the shape characteristics of traffic participants, and identify the type of object. The radar is installed at the front of the vehicle to detect kinematic information such as the relative distance and relative speed between the vehicle and the obstacles in front. The traffic participant information acquisition refers to detecting the movement trajectory, movement speed and obstacle type of obstacles around the vehicle. The characterization parameters include: veh (t) = {v veh ,P veh ,a veh , C veh }, where T veh (t) represents the position of the vehicle after time t, v self , a self are the vehicle speed and acceleration information, C veh Indicates the additional risk coefficients of obstacles with different characteristics. Traffic participants with different characteristics use a neural network model to process the information collected by the external camera to complete the identification and classification of potential collision targets in front. The main classification features include their size, color characteristics, etc. The additional risk coefficient of each type of target is subsequently determined based on its collision response characteristics; The prediction of vehicle collision risk in step 3 is to determine the collision risk based on the communication interval of the intersection of the vehicle and the obstacle trajectory, t col is the nearest collision risk time node; the current risk of the vehicle is determined by the minimum acceleration required for vehicle collision avoidance; the parameter R col Characterizes the size of the collision risk faced. This parameter can be obtained by the following equation: Among them, a x_n 、a y_n are the minimum longitudinal and lateral accelerations required to push the collision avoidance time point beyond the risk time threshold, R col is the vehicle collision risk parameter, F lane is the feasibility judgment of lane change, 0 means that lane change is not allowed, 1 means that lane change is allowed, a x_max , a y_max Represents the maximum longitudinal deceleration and lateral acceleration allowed by the current road, K x_n , K y_n They are the longitudinal deceleration risk factor and the lateral acceleration risk factor respectively; The occupant injury risk prediction module establishes a collision accident simulation database and then establishes an occupant injury risk prediction model based on machine learning. The occupant injury risk prediction model adopts a hybrid model (Hybrid MLP-SVM model) based on an improved multi-layer perceptron (MLP) and a support vector machine (SVM) to perform prediction tasks for high-dimensional input problems. The first level of the hybrid model is an MLP model, which is divided into an input layer, a hidden layer, and an output layer. The preset occupant injury influencing factors are used as the input of the model. The hidden layer uses the ReLU activation function to improve the model's ability to handle nonlinear problems. The feature vector input by the output layer is passed to the second-level SVM model, and the extracted features are used as the input of the SVM. The SVM is trained to perform classification or regression tasks. At the same time, the radial basis kernel function (RBF) is used to handle nonlinear separable situations, and finally the prediction results of occupant injury are obtained; The establishment process of the collision accident simulation database includes the following steps: first, establish a vehicle collision occupant injury value analysis model under the intervention of the active safety system, including the vehicle cockpit model, the occupant restraint system model and the human finite element model, reconstruct a typical accident scene on the simulation platform, and use the vehicle motion waveform generated during the collision avoidance process as the boundary condition for solving the occupant injury value analysis model, and analyze the kinematic response of the vehicle during the active safety system intervention process and the occupant injury evaluation index value; in order to ensure the rationality of the sample distribution in the data set, the vehicle motion state at the moment of collision, the action time and intensity of the active safety system, the use of the occupant seat belt, the occupant sitting posture, the seat back angle, and the front and rear displacement of the seat are used as variables. Based on the experimental design method, the occupant injury value analysis model under different parameters is simulated and solved, and the dynamic characteristics of the head, neck, chest and other parts of the dummy model are obtained, and an occupant response data set with a sufficient number of samples is established; among them, 80% of the sample data are used as training samples, and 20% of the sample data are used as test samples; due to the large number of influencing parameters in the experiment, the results are subjected to Pearson correlation analysis, and the calculation method is as follows, and the factors that have a greater impact on occupant injury are determined based on the calculation results; Among them, X i represents the above-mentioned influencing factors, Y i An evaluation index value representing occupant injury; Step 4 The collaborative control module selects the best driving control strategy in real time based on the information detected and calculated by each system; when the vehicle is in a typical risk scenario such as a blind spot with limited perception or an intersection with complex traffic conditions, the system coordination control module selects the best driving control strategy in real time based on the vehicle collision risk index R calculated in real time by the risk prediction module. col and occupant injury risk indicators AIS and WIC to formulate appropriate coordinated control strategies; when the vehicle collision risk index exceeds the set threshold, it is determined whether the vehicle's collision avoidance capability can avoid the accident under the current working conditions. If the accident can be avoided, the system will give priority to controlling the vehicle to automatically complete the collision avoidance; if the collision accident is unavoidable, the system will combine the restraint system information, occupant posture information, vehicle motion information and collision target information to formulate control rules for the vehicle's braking and steering systems, adjust the vehicle restraint system status, and reduce the AIS occupant injury risk to a low level; when the AIS score level cannot be lowered, the optimization goal is to reduce the WIC score.

Images