An auxiliary safety control method and system based on a driver behavior intention recognition model

Abstract

The application discloses an auxiliary safety control method based on a driver behavior model, comprising the following steps: establishing a driver behavior intention recognition model; acquiring spatial attribute values of target objects within a sensing range; calculating a current vehicle motion trajectory; screening target objects according to the spatial attribute values of the target objects, and screening out the closest target of interest in each region; calculating the relative motion relationship between the vehicle and the target of interest, and classifying the target of interest; calculating TTCT and TTC, and outputting a collision signal if TTC is less than TTCT; acquiring driver behavior characteristics and inputting the driver behavior characteristics into the driver behavior intention recognition model to obtain the behavior intention of the current driver; and deciding whether to output a terminal braking instruction according to the behavior intention; and outputting the terminal braking instruction to an execution terminal if there is no behavior intention of subjectively avoiding a collision. The application can effectively adapt to a driving test application scenario and ensure driving safety during a driving test.

Description

Technical Field

[0001] This invention relates to the fields of intelligent driver's license testing and intelligent driving of automobiles, specifically to a method and system for controlling an auxiliary safety system based on a driver behavior intention recognition model. Background Technology

[0002] Intelligent driving tests have been a hot topic in the driving test industry in recent years. The aim is to introduce existing advanced technologies, combined with the specific characteristics of driving test scenarios, to reduce human intervention in the testing process and ensure fairness and transparency. Currently, Advanced Driver Assistance Systems (ADAS) are widely used in road vehicles and have shown good performance in reducing road accidents. However, ADAS technology is currently mainly designed for experienced drivers and only provides assisted driving functions, and there are still instances of accidental braking or missed braking. Because drivers of driving test vehicles do not possess full driving ability and their driving behavior is immature, it is difficult to directly introduce existing ADAS technology into driving test vehicles operated by immature drivers.

[0003] Currently, driving safety during driving tests is primarily ensured through the intervention of a safety officer in the passenger seat. However, this process is prone to human intervention, including prompting and even cheating. While monitoring systems for safety officers exist, the sheer volume of data makes full-process monitoring difficult. Furthermore, the manpower required for safety officer monitoring results in significant operational costs. Therefore, it is necessary to develop an auxiliary safety product suitable for driving tests. This would not only effectively improve the fairness and transparency of driving tests but also significantly reduce labor costs. Summary of the Invention

[0004] To address the issue that existing automatic emergency braking technology cannot be adapted to driving test vehicles, this invention provides an automatic emergency braking control method and system based on a driver behavior intention recognition model. By introducing a driver behavior model to determine the terminal braking command, it can effectively adapt to driving test application scenarios and ensure driving safety during the driving test process, especially in the absence of an accompanying safety officer.

[0005] The first aspect of this invention discloses an auxiliary safety control method based on a driver's test behavior model, comprising:

[0006] A driver behavior intention recognition model is established based on a driver behavior habit database. The driver behavior habit database contains the correspondence between driver behavior characteristics and vehicle response data at corresponding times, and the vehicle response data is used to reflect the driver's behavioral intention.

[0007] The spatial attribute values ​​of target objects within the perception range are obtained through the environmental sensing module on the test vehicle;

[0008] Vehicle basic data is acquired via the CAN bus, and the current vehicle trajectory is calculated based on the vehicle basic data.

[0009] Based on the spatial attribute values ​​of the target objects, target objects are filtered to identify the targets of interest that are closest to the vehicle in each pre-divided area.

[0010] Calculate the relative motion relationship between the vehicle and the target based on the spatial attribute values ​​of the vehicle and the target, and classify the target.

[0011] Calculate the collision time threshold under the current vehicle trajectory, correct the collision time threshold according to the motion relationship between the vehicle and the target of interest, calculate the collision time between the target of interest and the vehicle, and if the collision time is less than the corrected collision time threshold, output the corresponding collision signal.

[0012] Upon receiving a collision signal, the driver's behavioral characteristics collected by the visual sensor are acquired, and the driver's behavioral characteristics are input into the driver's behavioral intent recognition model to obtain the current driver's behavioral intent.

[0013] The decision on whether to output a terminal braking command is based on the stated behavioral intent: if the driver does not have a subjective intent to avoid a collision, then a terminal braking command is output to the execution terminal; otherwise, no terminal braking command is output.

[0014] Furthermore, the auxiliary safety control method also includes: the execution terminal actively braking the vehicle after receiving a braking command.

[0015] An auxiliary safety control system based on a driving test behavior model is characterized by comprising a main control module, and a vision sensor, a vehicle response data acquisition module, an environmental perception module, and a vehicle basic data acquisition module electrically connected to the main control module.

[0016] The visual sensors include two sets installed on the top and bottom of the driver's cabin, respectively, to acquire the driver's hand and head postures and movements, as well as the driver's leg postures and movements.

[0017] The vehicle response data acquisition module is connected to the vehicle's OBD and is used to acquire vehicle dynamics response data;

[0018] The environmental perception module is used to dynamically perceive the road environment around the vehicle and output the spatial attribute values ​​of the target objects within the perception range.

[0019] The vehicle basic data acquisition module is connected to the CAN bus and uses the CAN communication network to perform data parsing and synchronization to acquire vehicle basic data.

[0020] The main control module includes a processor and a memory, the memory storing a computer program; the processor calls the computer program to execute the automatic emergency braking control method based on the driving test behavior model described in the first aspect of the present invention.

[0021] Furthermore, the main control module is also connected to an external braking actuator; after receiving the braking command sent by the main control module, the braking actuator performs a braking response.

[0022] This invention relates to an automatic emergency braking control method and system based on a driver behavior intent recognition model. By introducing a driver behavior model to determine the final braking command, it can effectively adapt to driving test application scenarios. Simultaneously, by screening targets in various areas around the vehicle and calculating collision times separately, it not only avoids frontal collisions but also achieves collision control for side and rear vehicles. It can essentially replace the role of the onboard safety officer, improving the intelligence and automation of driving tests, fundamentally resolving controversies surrounding premature safety officer supervision and intervention in driving tests, and further enhancing the transparency and fairness of the examination process. Attached Figure Description

[0023] Figure 1 This is a schematic diagram of the control method for an auxiliary safety system based on a driver behavior and intent recognition model.

[0024] Figure 2 This is a schematic diagram of environmental perception.

[0025] Figure 3 This is a schematic diagram of the region of interest.

[0026] Figure 4 This is a block diagram of an auxiliary safety control system based on a driver behavior intent recognition model. Detailed Implementation

[0027] To help you understand this invention, the technical solution of this invention will be further explained below with reference to specific embodiments.

[0028] Combination Figure 1 As shown in the figure, this invention discloses an auxiliary safety system control method based on a driver behavior intention recognition model, which mainly includes the following steps:

[0029] S1: Based on the driver behavior habit database, establish a driver behavior intention recognition model to identify the driver's behavioral intention.

[0030] Before establishing a driver behavior intention recognition model, a driver behavior habit database must first be created. The initial model is then trained using data from the database to obtain the desired driver behavior intention recognition model.

[0031] The driver behavior database contains the correspondence between driver behavior characteristics and the corresponding vehicle responses at different times. For example, when a driver presses the brake pedal, the vehicle speed decreases; when a driver turns the steering wheel to the left, the vehicle begins to change lanes to the left.

[0032] The driver's behavioral characteristics can be collected through visual sensors installed in the cockpit of the test vehicle. Specifically, visual sensors are installed at the top and bottom of the cockpit. The top visual sensor is mainly used to identify the driver's hand and head postures and movements, while the bottom visual sensor is mainly used to identify the driver's leg postures and movements. The collected behavioral characteristics mainly include hand grip on the steering wheel, accelerator pedal operation, brake pedal operation, clutch pedal operation, head rotation, and gaze shift. Furthermore, the collected driver behaviors can be pre-filtered to remove behaviors such as touching the head or scratching the ear, thus selecting the valid driver behavioral characteristics.

[0033] Vehicle response data, also known as vehicle dynamics response, can be obtained through the vehicle's OBD (On-Board Diagnostics) system. Specifically, the OBD system collects vehicle response data at corresponding moments based on driver behavior characteristics, thus establishing a correlation between different driver behaviors and vehicle responses. Vehicle response data mainly includes: vehicle acceleration, vehicle deceleration, changing lanes to the left, changing lanes to the right, turning left, turning right, and making U-turns.

[0034] After acquiring the two types of data, a one-to-one correspondence is established between valid driver behavior characteristics and vehicle response data, thereby obtaining a driver behavior habit database. The samples in the data database can be represented as X. i (A i B i ), where A i B represents the effective behavioral characteristics of the i-th driver. i This represents the vehicle dynamics response corresponding to the effective behavioral characteristics of the i-th driver.

[0035] Understandably, vehicle response data is usually an objective reflection of the driver's behavioral intentions. In other words, by collecting vehicle response data, the driver's behavioral intentions can be determined. Therefore, the driver's behavioral intentions can be mainly divided into several categories, such as acceleration, deceleration, left lane change, right lane change, left turn, right turn, and U-turn. Taking the accelerator pedal as an example, when the driver presses the accelerator pedal, the corresponding vehicle response data shows that the vehicle speed increases, indicating that the driver intends to accelerate; when the driver lifts their foot off the accelerator pedal, the corresponding vehicle response data shows that the vehicle speed decreases, indicating that the driver intends to decelerate. Taking the steering wheel as an example, when the driver's hands grip the steering wheel and turn it slightly to the left, the corresponding vehicle response data shows that the vehicle begins to change lanes to the left, indicating that the driver intends to change lanes.

[0036] Based on a pre-built database of driver behavior habits, a driver behavior intention recognition model is established to dynamically identify driver behavior intentions. The model establishment process mainly includes:

[0037] Construct an initial deep neural network (DNN) model, which includes an input layer, hidden layers, and an output layer. The hidden layer neuron model is y = f(ΣW). i *X i +b), where W i Here, b represents the corresponding weights, f is the bias term, and f is the activation function, where f(x) = sigmoid(x); the weights are updated to W. i =W i-1 -η*(ΔL / Δw), where η represents the learning rate, ΔL / Δw represents the gradient of the loss function relative to the weights, and L represents the total number of forward layers; obtain data samples X from the driver behavior habit database. i (A i B i The deep neural network is trained using the input data. After training, the required driver behavior intention recognition model is obtained. Based on this model, the corresponding vehicle dynamic response can be output after inputting the driver's behavior characteristics, thereby knowing the driver's behavior intention.

[0038] S2: Obtain the spatial attribute values ​​of the target objects within the perception range through the environmental sensing module on the test vehicle.

[0039] The environmental perception module is mainly used to dynamically perceive the road environment around the vehicle and output the spatial attribute values ​​of targets within the perception range. The spatial attribute values ​​mainly include location information, velocity information, acceleration information, category information, size information, orientation information, etc.

[0040] like Figure 2As shown, the environmental perception module installed on the test vehicle mainly includes millimeter-wave radar, cameras, and lidar. The lidar is installed on the top of the vehicle to collect point cloud data within a 360° range. Two cameras are installed on each side of the vehicle (front, rear, and sides) to fuse with the lidar point cloud data for target clustering and recognition. Millimeter-wave radars are installed at the four corners of the vehicle, primarily used to correct the speed of targets.

[0041] S3: Acquires basic vehicle data based on the CAN bus and calculates the current vehicle trajectory.

[0042] Before calculating the vehicle's trajectory, it is necessary to acquire basic vehicle data. This data can be obtained through a vehicle basic data acquisition module, which is connected to the CAN bus and uses the CAN communication network for data parsing and synchronization. The basic vehicle data mainly includes real-time data such as vehicle speed, gear position, steering wheel angle, yaw rate, braking, and throttle.

[0043] Once the basic vehicle data is acquired, the current vehicle trajectory can be calculated based on this data.

[0044] For example, the vehicle's trajectory is calculated as follows:

[0045] First, based on the steering wheel angle α, calculate the current front wheel angle θ:

[0046] θ = K * α, where K is the vehicle steering ratio;

[0047] Based on the obtained front wheel steering angle θ, the vehicle's turning radius R is then calculated:

[0048] R = vehicle wheelbase / sin(θ),

[0049] Therefore, the current vehicle trajectory is the turning radius R ± vehicle width * 0.5.

[0050] S4: Based on the spatial attribute values ​​of the target objects provided by the environmental perception module, filter the target objects and select the target of interest that is closest to the vehicle in each pre-divided area.

[0051] Before selecting target objects, the space around the vehicle is first divided into eight areas: front left, front directly, front right, rear left, rear directly, rear right, left side, and right side. Figure 3 As shown; then, based on the spatial attribute values ​​of the target object (mainly referring to location information and size information), the target object in each area is filtered to find the target object closest to the vehicle within each area and take it as the target of interest in the corresponding area.

[0052] S5: Calculate the relative motion relationship between the vehicle and the target based on the spatial attribute values ​​of the vehicle and the target, and classify the target.

[0053] Based on the spatial attribute values ​​of the target objects obtained by the environmental perception module, the relative motion relationship between the vehicle and each target of interest is calculated. The calculation method is as follows: the X-axis of the vehicle's planar coordinate system is defined to coincide with the vehicle's centerline, with the direction pointing towards the front of the vehicle being positive; the Y-axis is perpendicular to the X-axis, with the direction pointing to the driver's left being positive; the relative velocity Vr between the target of interest and the vehicle is defined as Vr = target of interest velocity – vehicle speed, with the direction of the relative velocity along the coordinate axis being positive.

[0054] Based on the relative speed and the direction of relative speed, that is, the relative motion relationship between the vehicle and the target, the targets of interest in each area are classified into categories such as stationary targets, targets moving away from each other, targets approaching at a constant speed, targets approaching at an accelerating speed, and targets approaching at a decelerating speed.

[0055] S6: Calculate the collision time threshold under the current vehicle trajectory, correct the collision time threshold according to the motion relationship between the target of interest and the vehicle, and then calculate the collision time between the target of interest and the vehicle. If the collision time is less than the corrected collision time threshold, it means that the basic triggering condition is met, and the corresponding collision signal is output.

[0056] Specifically, the TTCT0 (time to collision threshold) of the current vehicle trajectory is first calculated based on the current vehicle speed. The collision time threshold is proportional to the vehicle speed.

[0057] TTCT0 = (V + value) / factor

[0058] In the formula, V represents the current vehicle speed, value represents speed compensation, and factor represents a coefficient. In this embodiment of the invention, value = 80 and factor = 78.5 can be set.

[0059] Next, based on the classification of targets of interest in each region, targets approaching the vehicle (i.e., targets that may collide) are selected from each region. The collision time threshold for these targets in each region is then adjusted: TTCT0 is increased for accelerating, constant-speed, and stationary targets, and decreased for decelerating targets. For example, the adjustment can be performed using the following formula:

[0060] Stationary target: TTCT = TTCT0 + 0.2;

[0061] Accelerate towards the target: TTCT = (TTCT0 + 0.2) * (0.1794 + 0.026 * Vr);

[0062] Decelerate approaching the target. Target: TTCT = (TTCT0 + 0.2) * (0.11888 - 0.00456 * Vr);

[0063] Approach the target at a constant speed: TTCT = (TTCT0 + 0.2) * (0.095 + 0.024 * Vr);

[0064] In the formula, TTCT represents the corrected collision time threshold; Vr represents the relative speed between the target of interest and the vehicle.

[0065] Then, calculate TTC (time to collision), that is, based on the attribute values (speed, acceleration, and distance from the vehicle) of the target of interest in each area, calculate the time required for the vehicle to collide with the target of interest if the current motion relationship is maintained.

[0066] TTC = S / Vr,

[0067] In the formula, S represents the relative distance between the target of interest and the vehicle, and Vr represents the relative speed of the target of interest compared to the vehicle.

[0068] Finally, compare the calculated collision time TTC with the corrected collision time threshold TTCT: If the TTC of the target in the area < TTCT, then output the basic collision signal (B1).

[0069] S7: After receiving the basic collision signal, decide whether to output the terminal collision signal according to the classification result of the driver behavior intention recognized by the driver behavior intention recognition model: If the driver has no subjective intention to avoid collision, then output a braking instruction to the execution terminal, otherwise do not output a braking instruction.

[0070] After receiving the basic collision signal, obtain the driver behavior characteristics collected by the vision sensor, and input the driver behavior characteristics into the driver behavior intention recognition model to obtain the current driver's behavior intention.

[0071] The system makes a decision according to the classification result of the driver behavior intention: If the driver has no subjective intention to avoid collision, then output the corresponding braking instruction (B2) to the execution terminal; if the driver has a subjective intention to avoid collision, then suspend the output of the braking instruction.

[0072] For example, during a vehicle's straight - line driving, after receiving the basic collision signal in the front - directly - ahead area, the classification result output by the driver behavior model is that the driver has no subjective intention to avoid collision (for example, the driver still maintains or increases the current throttle pedal travel). At this time, a braking instruction should be output to the execution terminal to avoid an accident.

[0073] Upon receiving a braking command, the actuator actively brakes the vehicle. If it is determined that the driver has no subjective intention to avoid a collision, the actuator (i.e., the "braking actuator") will actively brake the vehicle after receiving the braking command (B2) issued by the system decision to avoid a collision and ensure the safety of the driving test process.

[0074] Understandably, this step is executed by an external actuator. As an optional method, the external braking mechanism mainly consists of a controller, motor, wire rope, and connectors. The controller communicates with the system via serial communication. When the controller receives a braking command, it controls the motor to rotate forward, retracting the wire rope, pressing down the brake pedal, and actively braking. When the controller receives a release command, it controls the motor to rotate in reverse, releasing the wire rope and freeing the brake pedal.

[0075] It is understood that the implementation steps of the auxiliary safety control method based on the driver behavior intention recognition model disclosed in the embodiments of the present invention are not completely fixed and can be adjusted according to actual conditions.

[0076] Combination Figure 4 As shown, another embodiment of the present invention discloses an automatic emergency braking control system based on a driver behavior intention recognition model, which mainly includes a main control module, and a visual sensor, a vehicle response data acquisition module, an environmental perception module, and a vehicle basic data acquisition module electrically connected to the main control module.

[0077] The visual sensors primarily consist of two sets mounted on the top and bottom of the driver's cabin, used to acquire the driver's hand and head postures and movements, and leg postures and movements, respectively. The vehicle response data acquisition module connects to the vehicle's OBD system to acquire vehicle dynamics response data. The environmental perception module is mainly used to dynamically perceive the road environment surrounding the vehicle and output spatial attribute values ​​of targets within the perception range. It typically consists of millimeter-wave radar, cameras, and lidar, and its workflow is as follows... Figure 2 The vehicle basic data acquisition module is connected to the CAN bus and uses the CAN communication network for data parsing and synchronization to acquire basic vehicle data. In addition, the main control module is also connected to the controller in the actuator to issue braking commands to the actuator, which then performs the braking action to avoid a collision. The main control module includes a processor and a memory. The memory stores a computer program for implementing the automatic emergency braking control method based on the driver's behavioral intent recognition model described in the above embodiments. The processor calls the computer program in the memory to execute the automatic emergency braking control method. The functional characteristics and specific implementation methods of the above modules can be found in the foregoing method embodiments, and will not be repeated here.

[0078] Finally, it should be noted that although the embodiments of the present invention have been described above in conjunction with the accompanying drawings, the present invention is not limited to the specific embodiments and application fields described above. The specific embodiments described above are merely illustrative and instructive, and not restrictive. Those skilled in the art, guided by this specification, can make many other forms without departing from the scope of protection of the claims of the present invention, and all of these are within the scope of protection of the present invention.

Claims

1. An auxiliary safety control method based on a driver's test behavior model, characterized in that, include: A driver behavior intention recognition model is established based on a driver behavior habit database. The driver behavior habit database contains the correspondence between driver behavior characteristics and vehicle response data at corresponding times, and the vehicle response data is used to reflect the driver's behavioral intention. The spatial attribute values ​​of target objects within the sensing range are obtained through the environmental perception module on the test vehicle; Vehicle basic data is acquired via the CAN bus, and the current vehicle trajectory is calculated based on the vehicle basic data. Based on the spatial attribute values ​​of the target objects, target objects are filtered to identify the targets of interest that are closest to the vehicle in each pre-divided area. Calculate the relative motion relationship between the vehicle and the target based on the spatial attribute values ​​of the vehicle and the target, and classify the target. Calculate the collision time threshold under the current vehicle trajectory, correct the collision time threshold according to the motion relationship between the vehicle and the target of interest, calculate the collision time between the target of interest and the vehicle, and if the collision time is less than the corrected collision time threshold, output the corresponding collision signal. Upon receiving a collision signal, the driver's behavioral characteristics collected by the visual sensor are acquired, and the driver's behavioral characteristics are input into the driver's behavioral intent recognition model to obtain the current driver's behavioral intent. The decision on whether to output a terminal braking command is based on the stated behavioral intent: if the driver does not have a subjective intention to avoid a collision, then an terminal braking command is output to the execution terminal; otherwise, no terminal braking command is output. The step of establishing a driver behavior intention recognition model based on a driver behavior habit database includes: The driver's behavioral characteristics are collected by visual sensors installed in the cockpit of the test vehicle, and then filtered to obtain effective driver behavioral characteristics. Vehicle response data, also known as vehicle dynamics response, is obtained through vehicle OBD. Establish an effective one-to-one correspondence between driver behavior characteristics and vehicle response data to obtain a database of driver behavior habits. The data sample is represented as X. i (A) i B i ), where A i B represents the effective behavioral characteristics of the i-th driver. i This represents the vehicle dynamics response corresponding to the effective behavioral characteristics of the i-th driver; Construct an initial deep network model, which includes an input layer, hidden layers, and an output layer. The hidden layer neuron model is y = f(ΣW). i *X i + b), where W i Here, b represents the corresponding weights, f is the bias term, and f is the activation function, where f(x) = sigmoid(x); the weights are updated to W. i =W i-1 -η*(ΔL / Δw), where η represents the learning rate, ΔL / Δw represents the gradient of the loss function relative to the weights, and L represents the total number of forward layers; the data samples X from the driver behavior habit database... i (A) i B i The deep neural network is trained using the data as training data; once training is complete, the desired driver behavior and intention recognition model is obtained. The process of calculating the relative motion relationship between the vehicle and the target of interest based on their spatial attribute values, and classifying the target of interest, specifically includes: Based on the spatial attribute values ​​of the target objects obtained by the environmental perception module, the relative motion relationship between the vehicle and each target of interest is calculated as follows: Define the vehicle's planar coordinate system such that the X-axis coincides with the vehicle's centerline, pointing towards the front of the vehicle as positive, and the Y-axis is perpendicular to the X-axis, pointing towards the driver's left as positive. Define the relative velocity Vr between the target of interest and the vehicle as Vr = target of interest velocity – vehicle speed, with the direction of relative velocity being positive along the coordinate axis; Based on the relative speed and the direction of the relative speed, that is, the relative motion relationship between the vehicle and the target of interest, the targets of interest in each area are classified. The classification categories include stationary targets, targets moving away from each other, targets approaching at a constant speed, targets approaching at an accelerating speed, and targets approaching at a decelerating speed.

2. The auxiliary safety control method as described in claim 1, characterized in that, The driver's behavioral characteristics include characteristics of hand gripping the steering wheel, accelerator pedal operation, brake pedal operation, clutch pedal operation, head turning, and gaze shifting. The vehicle response data includes: vehicle acceleration, vehicle deceleration, vehicle changing lanes to the left, vehicle changing lanes to the right, vehicle turning left, vehicle turning right, and vehicle making a U-turn; correspondingly, the driver's behavioral intentions include several categories such as acceleration, deceleration, changing lanes to the left, changing lanes to the right, turning left, turning right, and making a U-turn.

3. The auxiliary safety control method as described in claim 1 or 2, characterized in that, The spatial attribute values ​​include location information, speed information, acceleration information, size information, and orientation information; the vehicle basic data includes vehicle speed, gear, steering wheel angle, yaw rate, braking, and throttle.

4. The auxiliary safety control method as described in claim 1 or 2, characterized in that, Target objects are filtered based on their spatial attribute values, identifying the closest targets of interest to the vehicle within pre-divided regions. Specifically, this includes: Before screening the target object, the space around the vehicle is divided into eight areas: front left, front right, front left, rear left, rear right, left side, and right side. Then, based on the spatial attribute values ​​of the target objects, target objects are filtered in each region to find the target objects closest to the vehicle within each region and these are selected as targets of interest in the corresponding region.

5. The auxiliary safety control method as described in claim 1 or 2, characterized in that, The formula for calculating the collision time threshold is: TTCT0 = (V + value) / factor In the formula, TTCT0 represents the collision time threshold, V represents the current vehicle speed, value is the speed compensation, and factor is the coefficient; The formula for calculating the collision time is: TTC = S / Vr In the formula, S represents the relative distance between the target of interest and the vehicle, and Vr represents the relative speed between the target of interest and the vehicle.

6. The auxiliary safety control method as described in claim 5, characterized in that, The collision time threshold is corrected based on the motion relationship between the vehicle and the target of interest, including: Stationary target: TTCT = TTCT0 + 0.2; Accelerate towards the target: TTCT = (TTCT0 + 0.2) * (0.1794 + 0.026 * Vr); Decelerate as you approach the target: TTCT = (TTCT0 + 0.2) * (0.11888 - 0.00456 * Vr); Approaching the target at a constant speed: TTCT = (TTCT0 + 0.2) * (0.095 + 0.024 * Vr); In the formula, TTCT represents the corrected collision time threshold; Vr represents the relative speed between the target of interest and the vehicle.

7. The auxiliary safety control method as described in claim 1 or 2, characterized in that, The vehicle's trajectory is calculated as follows: Firstly, based on the steering wheel angle α Calculate the current front wheel steering angle θ : θ = K * α ,in K The vehicle's steering ratio; Based on the obtained front wheel steering angle θ Calculate the vehicle's turning radius R : R = Vehicle wheelbase / sin( θ ), Therefore, the current vehicle trajectory is: turning radius R ±vehicle width * 0.

5.

8. An auxiliary safety control system based on a driving test behavior model, characterized in that, It includes a main control module, as well as a vision sensor, a vehicle response data acquisition module, an environmental perception module, and a vehicle basic data acquisition module that are electrically connected to the main control module; The visual sensors include two sets installed on the top and bottom of the driver's cabin, respectively, to acquire the driver's hand and head postures and movements, as well as the driver's leg postures and movements. The vehicle response data acquisition module is connected to the vehicle's OBD and is used to acquire vehicle dynamics response data; The environmental perception module is used to dynamically perceive the road environment around the vehicle and output the spatial attribute values ​​of the target objects within the perception range. The vehicle basic data acquisition module is connected to the CAN bus and uses the CAN communication network to perform data parsing and synchronization to acquire vehicle basic data. The main control module includes a processor and a memory, the memory storing a computer program; the processor calls the computer program to execute the auxiliary safety control method based on the driving test behavior model as described in any one of claims 1 to 7.

Images