Method and device for preventing vehicle collision, computer equipment and storage medium

Abstract

The invention discloses a method and device for preventing vehicle collision, computer equipment and a storage medium, and the method comprises the steps: when a collision risk value between a vehicle and a target obstacle is greater than or equal to a preset collision risk threshold value, according to the actual steering operation torque of a driver at the current moment of the vehicle and a collision risk parameter, according to the method, the first reverse rotation torque at the next moment is predicted, when the next moment arrives, the actual steering operation torque of the driver at the next moment is completely counteracted through the first reverse rotation torque, and the AES function of the vehicle is activated. As the first reverse rotation torque can completely offset the actual steering operation torque of the driver at the next moment, the driver cannot operate the steering wheel, the problem that when the next moment arrives, the driver operates the steering wheel, so that the AES function cannot be activated, and collision accidents occur can be solved, vehicle collision is effectively prevented, and the safety of the vehicle is improved. And the safety of the vehicle in the running process is improved.

Description

Technical Field

[0001] The present application relates to the technical field of vehicle driving safety, and in particular to a method, device, computer equipment and storage medium for preventing vehicle collision. Background Art

[0002] With the continuous improvement of vehicle intelligent driving technology, users are paying more attention to the safety of vehicles during driving. When the vehicle is driving, when the vehicle system detects that the collision risk value of the vehicle reaches the preset collision risk threshold, the AES (Automatic Emergency Steering) function can be activated. If the driver turns the steering wheel during the activation of the AES function, the AES function will not be successfully activated, which may easily cause the vehicle to collide with obstacles. Summary of the invention

[0003] Based on this, a method, device, computer equipment and storage medium for preventing vehicle collision are provided to solve the problem in the prior art that the AES function cannot be activated due to the driver touching the steering wheel, thereby causing the vehicle to collide.

[0004] In a first aspect, a method for preventing a vehicle collision is provided, the method comprising:

[0005] Acquiring the actual steering operation torque of the driver of the vehicle at the current moment, and the collision risk parameter between the vehicle and the target obstacle at the current moment;

[0006] When the collision risk value between the vehicle and the target obstacle is greater than or equal to a preset collision risk threshold, predicting a first reverse rotation torque for completely offsetting the actual steering operation torque of the driver at a next moment according to the actual steering operation torque of the driver at the current moment and the collision risk parameter;

[0007] When the next moment arrives, the driver's actual steering operation torque at the next moment is completely offset by the first reverse rotation torque, and the AES function of the vehicle is activated.

[0008] Through the above method, when the collision risk value between the vehicle and the target obstacle is greater than or equal to the preset collision risk threshold, the first reverse rotation torque at the next moment is predicted based on the actual steering operation torque of the driver at the current moment and the collision risk parameter, thereby ensuring that the driver's steering wheel manipulation can be offset based on the first reverse rotation torque, thereby solving the problem of the AES function being unable to be activated due to the driver touching the steering wheel, thereby improving the safety of the vehicle during driving.

[0009] In one embodiment, when the next moment arrives, the first reverse rotation torque completely offsets the actual steering operation torque of the driver at the next moment, including:

[0010] When the next moment arrives, the actual steering operation torque of the driver at the next moment is converted to obtain the corresponding motor steering output torque; the first reverse rotation torque is greater than or equal to the motor steering output torque;

[0011] When the first reverse rotation torque is greater than the motor steering output torque, controlling the vehicle to steer in the direction of the first reverse rotation torque based on a difference between the first reverse rotation torque and the motor steering output torque;

[0012] When the first reverse rotation torque is equal to the motor steering output torque, the driving direction of the vehicle is maintained consistent with the driving direction at the last moment.

[0013] Through the above method, the actual steering operation torque of the driver at the next moment is first converted and processed to obtain the corresponding motor steering output torque, and then the magnitude of the first reverse rotation torque and the motor steering output torque are judged, and different measures are taken according to the judgment result to ensure that the vehicle can drive normally during the activation of the AES function, and when the first reverse rotation torque is greater than the motor steering output torque, the vehicle is controlled to turn in the direction of the first reverse rotation torque, which can better avoid obstacles and improve the safety of the vehicle during driving.

[0014] In one embodiment, the collision risk parameter includes at least one of an obstacle type parameter of the target obstacle, a behavior intention parameter of the target obstacle, an accelerator pedal parameter of the vehicle, a projection overlap ratio between the target obstacle and the vehicle, a relative distance between the vehicle and the target obstacle, and a relative speed between the vehicle and the target obstacle;

[0015] The predicting, based on the actual steering operation torque of the driver at the current moment and the collision risk parameter, a first reverse rotation torque for completely offsetting the actual steering operation torque of the driver at the next moment comprises:

[0016] The actual steering operation torque of the driver at the current moment and the collision risk parameter are input into the trained reverse rotation torque prediction model to obtain the first reverse rotation torque.

[0017] By using the above method, the first reverse rotation torque is predicted based on the trained reverse rotation torque prediction model, thereby improving the accuracy of the prediction result.

[0018] In one embodiment, before inputting the actual steering operation torque of the driver at the current moment and the collision risk parameter into the trained reverse rotation torque prediction model, the method includes:

[0019] Acquire a training data set and an initial model; the training data set includes the actual steering operation torque of the driver of the vehicle at multiple moments, the collision risk parameter between the vehicle and the target obstacle at each moment, and the actual steering operation torque of the driver of the vehicle at the next moment of each moment;

[0020] Performing training based on the training data set and the initial model;

[0021] When the preset training stop condition is reached, the reverse rotation torque prediction model is obtained.

[0022] Through the above method, training is performed based on the training data set and the initial model to obtain a reverse rotation torque prediction model to ensure the accuracy of the model training process.

[0023] In one embodiment, the reverse rotation torque predicted by the reverse rotation torque prediction model is positively correlated with a collision risk value between the vehicle and the target obstacle.

[0024] Through the above method, when the collision risk between the vehicle and the target obstacle is high, the corresponding reverse rotation torque is large, which can offset the driver's actual operating torque, ensuring that the driver's actual operating torque will not affect the activation of the AES function when the collision risk is high; when the collision risk between the vehicle and the target obstacle is low, the corresponding reverse rotation torque is small, and the driver can easily control the driving direction of the vehicle.

[0025] In one embodiment, the method further comprises:

[0026] When the collision risk value between the vehicle and the target obstacle is less than the preset collision risk threshold, the reverse rotation torque prediction model predicts a second reverse rotation torque according to the actual steering operation torque of the driver at the current moment and the collision risk parameter;

[0027] When the next moment arrives, the actual steering operation torque of the driver at the next moment is converted to obtain the corresponding motor steering output torque; the second reverse rotation torque is less than the motor steering output torque;

[0028] The vehicle is controlled to steer in the direction of the motor steering output torque according to a difference between the motor steering output torque and the second reverse rotation torque.

[0029] In the above method, when the collision risk between the vehicle and the target obstacle is low, the predicted second reverse rotation torque is less than the motor steering output torque corresponding to the driver's actual steering operation torque at the next moment, which can ensure that the driver controls the driving direction when the collision risk is low.

[0030] In one embodiment, the collision risk parameter includes at least two of an obstacle type parameter of the target obstacle, a behavior intention parameter of the target obstacle, an accelerator pedal parameter of the vehicle, a projection overlap rate between the target obstacle and the vehicle, a relative distance between the vehicle and the target obstacle, and a relative speed between the vehicle and the target obstacle; the collision risk value is obtained by:

[0031] Obtaining a weight coefficient corresponding to each of the collision risk parameters;

[0032] Based on the collision risk parameters and the weight coefficients, a collision risk value between the vehicle and the target obstacle is calculated.

[0033] In the above method, the collision risk value is calculated by using the collision risk parameter and the corresponding weight coefficient, thereby ensuring the accuracy of the collision risk calculation result.

[0034] In a second aspect, the present application provides a device for preventing vehicle collision, the device comprising:

[0035] An acquisition module, used to acquire the actual steering operation torque of the driver of the vehicle at the current moment, and a collision risk parameter between the vehicle and the target obstacle at the current moment;

[0036] a prediction module, configured to predict, when a collision risk value between the vehicle and the target obstacle is greater than or equal to a preset collision risk threshold, a first reverse rotation torque for completely offsetting the actual steering operation torque of the driver at a next moment according to the actual steering operation torque of the driver at the current moment and the collision risk parameter;

[0037] The control module is used for completely offsetting the actual steering operation torque of the driver at the next moment by the first reverse rotation torque when the next moment arrives, and activating the AES function of the vehicle.

[0038] In a third aspect, the present application provides a computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the method for preventing vehicle collision of the first aspect when executing the computer program.

[0039] In a fourth aspect, the present application provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the method for preventing vehicle collision of the first aspect described above.

[0040] The above-mentioned method, device, computer equipment and storage medium for preventing vehicle collision, when the collision risk value between the vehicle and the target obstacle is greater than or equal to the preset collision risk threshold, predicts a first reverse rotation torque for completely offsetting the driver's actual steering operation torque at the next moment based on the driver's actual steering operation torque at the current moment of the vehicle and the collision risk parameter. When the next moment arrives, the driver's actual steering operation torque at the next moment is completely offset by the first reverse rotation torque, and the vehicle's AES function is activated. Through the above method, when the next moment arrives, the driver cannot successfully operate the steering wheel because the first reverse rotation torque can completely offset the driver's actual steering operation torque at the next moment. Therefore, the problem of the AES function being unable to be activated and causing a collision accident due to the driver's operation of the steering wheel when the next moment arrives can be solved, which effectively prevents the vehicle from colliding and improves the safety of the vehicle during driving. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 A diagram of an application environment of a method for preventing vehicle collision in one embodiment;

[0042] Figure 2 A schematic flow chart of steps of a method for preventing vehicle collision in one embodiment;

[0043] Figure 3 A schematic diagram of controlling vehicle steering based on actual motor output torque in a vehicle system in an embodiment;

[0044] Figure 4 A schematic diagram of a process for training a prediction model in one embodiment;

[0045] Figure 5 A logic flow chart for preventing vehicle collision in one embodiment;

[0046] Figure 6 A schematic diagram of an AES scenario operation in an embodiment;

[0047] Figure 7 is a structural block diagram of a device for preventing vehicle collision in one embodiment;

[0048] Figure 8 FIG. 4 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0049] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0050] The present application provides a method for preventing vehicle collision, which can be applied to Figure 1 In the application environment shown. Among them, the terminal 102 communicates with the server 104 through the network. The terminal 102 can be a whole vehicle system in the vehicle. The whole vehicle system detects the collision risk parameters of the vehicle and the actual steering operation torque of the driver in real time, and reports the collision risk parameters and the actual steering operation torque of the driver to the server 104. The server 104 determines the reverse rotation torque according to all the collision risk parameters and the actual steering operation torque of the driver, and sends the reverse rotation torque to the whole vehicle system. Among them, the terminal 102 can be but not limited to various personal computers, whole vehicle systems, laptops, smart phones, tablet computers and portable wearable devices, and the server 104 can be implemented with an independent server or a server cluster composed of multiple servers. In some other embodiments, the terminal 102 can also determine the reverse rotation torque according to the collision risk parameters and the actual steering operation torque of the driver. The reverse rotation torque here includes a first reverse rotation torque and / or a second reverse rotation torque, and the difference between the first reverse rotation torque and the second reverse rotation torque is introduced below.

[0051] In previous technologies, when the vehicle is driving and the vehicle system detects that the collision risk value of the vehicle reaches a preset risk threshold, the AES function will be activated, causing the vehicle to perform emergency steering and / or emergency braking to avoid obstacles ahead. However, if the driver turns the steering wheel while the AES function is activated, the AES function will not be activated normally.

[0052] In view of this, if Figure 2 As shown, a method for preventing a vehicle from colliding with an obstacle is provided, comprising the following steps:

[0053] S201: Acquire the actual steering operation torque of the driver at the current moment of the vehicle, and the collision risk parameter between the vehicle and the target obstacle at the current moment.

[0054] S202: When the collision risk value between the vehicle and the target obstacle is greater than or equal to a preset collision risk threshold, a first reverse rotation torque for completely offsetting the driver's actual steering operation torque at the next moment is predicted based on the driver's actual steering operation torque at the current moment and the collision risk parameter.

[0055] S203: When the next moment arrives, the driver's actual steering operation torque at the next moment is completely offset by the first reverse rotation torque, and the AES function of the vehicle is activated.

[0056] The preset collision risk threshold in the embodiment of the present application is the threshold for triggering the activation of the AES function, that is, when the collision risk value between the vehicle and the target obstacle is greater than the preset collision risk threshold, the vehicle can be triggered to perform the operation of activating the AES function, and the specific size of this value can be flexibly set by the developer. In order to prevent the AES function from failing to activate due to the driver operating the steering wheel during the process of activating the AES function, the embodiment of the present application predicts in advance the first reverse rotation torque used to completely offset the actual steering operation torque of the driver at the next moment, so that when the next moment arrives, it is difficult for the driver to turn the steering wheel to ensure that the AES function can be successfully activated.

[0057] When the next moment arrives, the EPS (Electric Power Steering) in the vehicle system is used to process various signals in the vehicle steering process and control the output of the electric power motor. The electric power motor includes: a power steering module and a motor control module. When the next moment arrives, the actual steering operation torque of the driver at the next moment is converted and processed by the power steering module to obtain the corresponding motor steering output torque. The motor control module outputs the predicted first reverse rotation torque. The EPS calculates based on the motor steering output torque and the first reverse rotation torque to obtain the final actual motor output torque, and controls the steering of the vehicle based on the actual motor output torque.

[0058] It can be understood that as long as the first reverse rotation torque predicted in step S202 is greater than or equal to the motor steering output torque corresponding to the driver's actual steering operation torque at the next moment, the driver's actual steering operation torque at the next moment can be completely offset.

[0059] Preferably, the first reverse rotation torque is equal to the motor steering output torque corresponding to the actual steering operation torque of the driver at the next moment, so that when the next moment arrives, safe driving can be ensured at the minimum cost.

[0060] The schematic diagram of controlling vehicle steering based on actual motor output torque in the vehicle system is as follows: Figure 3As shown, the driver's actual steering operation torque is read by the hand torque sensor, and the hand torque sensor inputs the driver's actual steering operation torque into the power steering module in the EPS. The power steering module amplifies the driver's actual steering operation torque into a motor steering output torque. The collision risk value can be determined by the ADS (Adaptive amping System) or by a server that can communicate with the vehicle. Therefore, the hand torque sensor can also output the driver's actual steering operation torque to the ADS or the server, and the ADS or the server calculates the reverse rotation torque based on the collision risk value and the driver's actual steering operation torque, and sends the reverse rotation torque to the motor control module in the EPS. The EPS reads the motor steering output torque and the reverse rotation torque, calculates the actual motor output torque, and sends the actual motor output torque to the power steering motor, and the power steering motor controls the steering of the vehicle.

[0061] In the first example, when the collision risk between the vehicle and the target obstacle is small, for example, when the collision risk value between the vehicle and the target obstacle is less than a preset collision risk threshold, the AES function may not be activated, and there is no need to predict the reverse rotation torque at this time, that is, the operation of predicting the reverse rotation torque may not be performed at this time. When it is determined that the collision risk value between the vehicle and the target obstacle is greater than or equal to the preset collision risk threshold, the first reverse rotation torque used to completely offset the actual steering operation torque of the driver at the next moment may be predicted. In this way, there is no need to predict the reverse rotation torque in real time, which can improve the operating efficiency of the system.

[0062] It should be noted that the preset collision risk threshold in this example can be flexibly set by the developer according to the actual situation. In this example, the collision risk value between the vehicle and the target obstacle can be determined based on the collision risk parameters. The collision risk parameters in this example include but are not limited to the obstacle type parameters of the target obstacle, the behavior intention parameters of the target obstacle, the accelerator pedal parameters of the vehicle, the projection overlap rate between the target obstacle and the vehicle, the relative distance between the vehicle and the target obstacle, and at least two of the relative speed between the vehicle and the target obstacle. Specifically, the collision risk value can be determined in the following way:

[0063] Step 1: Obtaining a weight coefficient corresponding to each of the collision risk parameters;

[0064] Step 2: Based on each of the collision risk parameters and each of the weight coefficients, a collision risk value between the vehicle and the target obstacle is calculated.

[0065] The weight coefficient corresponding to each collision risk parameter may be pre-set in the vehicle, and there may be multiple weight coefficients corresponding to each collision risk parameter. For example, for a certain collision risk parameter, the correspondence between its collision risk parameter value range and the weight coefficient may be pre-set. When executing the above step one, the corresponding weight coefficient may be determined according to the collision risk parameter value range to which the collision risk parameter belongs.

[0066] In the above step 2, the collision risk value can be calculated according to the following formula:

[0067] A=OB_Typ*ω1+OB_Dst*ω2+OB_Spd*ω3+OB_Overlap*ω4+OB_Prepla

[0068] n*ω5+VCU_accPeActPos*ω6;

[0069] Among them, A is the collision risk value, OB_Typ is the obstacle type parameter, OB_Dst is the relative distance between the vehicle and the target obstacle, OB_Spd is the relative speed between the vehicle and the target obstacle, OB_Overlap is the projection overlap rate, OB_Preplan is the behavior intention parameter, VCU_accPeActPos is the accelerator pedal parameter, and ω1 to ω6 are the weight coefficients corresponding to each collision risk parameter.

[0070] In the second example, the reverse rotation torque at the next moment can be predicted in real time, where the reverse rotation torque includes a first reverse rotation torque and a second reverse rotation torque. When the next moment arrives, the driving direction of the vehicle is controlled based on the reverse rotation torque and the actual steering operation torque of the driver at the next moment. Specifically, when the collision risk value between the vehicle and the target obstacle is greater than or equal to the preset collision risk threshold, the first reverse rotation torque used to completely offset the actual steering operation torque of the driver at the next moment is predicted based on the actual steering operation torque of the driver at the current moment and the collision risk parameter, and when the collision risk value between the vehicle and the target obstacle is less than the preset collision risk threshold, the second reverse rotation torque that cannot completely offset the actual steering operation torque of the driver at the next moment is predicted.

[0071] When the collision risk value between the vehicle and the target obstacle is greater than or equal to the preset collision risk threshold, the predicted reverse rotation torque is the first reverse rotation torque, which can completely offset the driver's actual steering operation torque at the next moment. When the collision risk value between the vehicle and the target obstacle is less than the preset collision risk threshold, the predicted reverse rotation torque is the second reverse rotation torque, which cannot completely offset the driver's actual steering operation torque at the next moment.

[0072] Specifically, when the collision risk value between the vehicle and the target obstacle is greater than or equal to a preset collision risk threshold, a first reverse rotation torque used to completely offset the driver's actual steering operation torque at the next moment is predicted based on the driver's actual steering operation torque at the current moment and the collision risk parameter. When the next moment arrives, the driver's actual steering operation torque at the next moment is converted and processed to obtain the corresponding motor steering output torque; the first reverse rotation torque is greater than or equal to the motor steering output torque corresponding to the driver's actual steering operation torque at the next moment; when the first reverse rotation torque is greater than the motor steering output torque, the vehicle is controlled to turn in the direction of the first reverse rotation torque based on the difference between the first reverse rotation torque and the motor steering output torque; when the first reverse rotation torque is equal to the motor steering output torque, the driving direction of the vehicle is maintained consistent with the driving direction at the previous moment.

[0073] When the collision risk value between the vehicle and the target obstacle is less than a preset collision risk threshold, a second reverse rotation torque for partially offsetting the driver's actual steering operation torque at the next moment is predicted based on the driver's actual steering operation torque at the current moment and the collision risk parameter; when the next moment arrives, the driver's actual steering operation torque at the next moment is converted to obtain the corresponding motor steering output torque; the second reverse rotation torque is less than the motor steering output torque; and the vehicle is controlled to steer in the direction of the motor steering output torque based on the difference between the motor steering output torque and the second reverse rotation torque.

[0074] In this example, when the collision risk value between the vehicle and the target obstacle is greater than or equal to the preset collision risk threshold, the predicted first reverse rotation torque is greater than or equal to the motor steering output torque corresponding to the driver's actual steering operation torque at the next moment. This ensures that when the collision risk is large, the driver's actual operation torque at the next moment can be completely offset by the predicted first reverse rotation torque to avoid the AES function from being unable to activate due to driver interference.

[0075] When the collision risk value between the vehicle and the target obstacle is less than the preset collision risk threshold, the predicted second reverse rotation torque is less than the motor steering output torque corresponding to the driver's actual steering operation torque at the next moment. This ensures that when the collision risk is small, the driver can control the steering wheel by himself and give the driving authority to the driver. At this time, the driver can control the steering of the vehicle through the steering wheel.

[0076] In a possible design, the obstacle type parameter of the target obstacle can be obtained according to the obstacle type of the target obstacle, specifically in the following manner:

[0077] A target obstacle type to which the target obstacle belongs is determined, and a first obstacle type parameter corresponding to the target obstacle type is determined according to a preset first corresponding relationship, wherein the first corresponding relationship records each preset obstacle type and each first obstacle type parameter corresponding to the obstacle type. Therefore, the target obstacle type is matched with each preset obstacle type, an obstacle type consistent with the target obstacle type and a first obstacle type parameter corresponding to the obstacle type consistent with the target obstacle are determined, and the first obstacle type parameter is determined as the obstacle type parameter corresponding to the target obstacle.

[0078] Exemplarily, the obstacle types recorded in the first corresponding relationship include but are not limited to road signs, vehicles, pedestrians, etc. The obstacle types can be divided according to actual conditions, which will not be elaborated in detail here.

[0079] Optionally, the target obstacle type to which the target obstacle belongs may be a whitelist obstacle type or a general list obstacle type.

[0080] In one example, a whitelist obstacle list and a general list obstacle list may be pre-stored on the vehicle. The developer may pre-write obstacles that are more likely to collide with the vehicle into the general list obstacle list, such as children into the general list obstacle list, and obstacles that are less likely to collide with the vehicle into the whitelist obstacle list, such as adults into the whitelist obstacle list.

[0081] In another example, the vehicle can identify obstacles in the current driving path, and automatically generate a white list obstacle list and a general list obstacle list based on the identified obstacles. The target obstacle can be the first obstacle that the vehicle encounters in the future that is identified by the vehicle at the current moment.

[0082] Among them, the first risk value of the obstacles recorded in the white list obstacle list is lower than the second risk value of the obstacles recorded in the general list obstacle list. When the risk value of the obstacle is higher, the probability of the vehicle colliding with the obstacle is higher, that is, the obstacles in the general list obstacle list have a higher probability of colliding with the vehicle than the obstacles in the white list obstacle list.

[0083] The whitelist obstacle list can be shown in Table 1 below:

[0084] Table 1

[0085] Obstacle Name Perception distance Cone barrel 49 Water Horse 121 Triangle 102 Anti-collision barrel 98 vehicle 200 pedestrian 100 …… ……

[0086] In the above Table 1, various obstacles and the perceived distance of the vehicle to each obstacle are recorded. The perceived distance is obtained through the sensor in the vehicle. The various obstacles in the above Table 1 are only used as examples and will not be described in detail here.

[0087] The general list of obstacles is shown in Table 2 below:

[0088] Table 2

[0089] Obstacle Name Perception distance Fallen dummy 70 Balloon car rolls over 129 Mixed scene of stationary people and vehicles 102 Fallen two-wheeler 77 Carton(60*40) 80 Carton(80*50) 86 …… ……

[0090] In the above Table 2, various obstacles and the perceived distance of the vehicle to each obstacle are recorded. The various obstacles in the above Table 2 are only used as examples and will not be described in detail here.

[0091] In a possible design, the obstacle type parameter of the target obstacle can also be obtained in the following way:

[0092] The target obstacle is matched with each preset obstacle, each preset obstacle corresponds to a second obstacle type parameter, therefore, the preset obstacle consistent with the target obstacle is determined according to the corresponding relationship, and the second obstacle type parameter corresponding to the preset obstacle consistent with the target obstacle is determined, and the second obstacle type parameter is determined as the obstacle type parameter corresponding to the target obstacle.

[0093] In a possible design, the accelerator pedal parameter is determined based on the accelerator pedal opening of the vehicle. Different accelerator pedal openings can be set to correspond to different accelerator pedal parameters. The larger the accelerator pedal parameter, the more likely the vehicle is to collide with the target obstacle. For details, see the following Table 3:

[0094] Table 3

[0095] Accelerator pedal opening Accelerator pedal parameters 25％ 12 50％ 23 75％ 34 …… ……

[0096] In the above Table 3, different accelerator pedal openings correspond to different accelerator pedal parameters. The above accelerator pedal openings and accelerator pedal parameters can be adjusted according to actual conditions. The above Table 3 is only used as an example.

[0097] The projection overlap rate between the target obstacle and the vehicle represents the proportion of the overlapping area when the target obstacle and the vehicle are projected. The larger the projection overlap rate, the higher the risk of collision between the vehicle and the target obstacle.

[0098] In one example, the projection overlap rate between the target obstacle and the vehicle can be determined based on the lateral projection area of ​​the vehicle and the area of ​​the target obstacle recognized by the vehicle, and the projection overlap rate between the target obstacle and the vehicle can be calculated based on the formula C1=S1 / (S1+S2), where C1 represents the projection overlap rate, S1 represents the area of ​​the target obstacle recognized by the vehicle, and S2 represents the lateral projection area of ​​the vehicle. For example, if the lateral projection area of ​​the vehicle is 12, and the area of ​​the target obstacle recognized by the vehicle is 18, the projection overlap rate of the target obstacle is calculated to be 18 / (12+18)=60%.

[0099] In another example, image acquisition may be performed on the target obstacle to obtain a target image, and the projection overlap rate may be calculated based on an area of ​​the target obstacle in the target image and an area of ​​the target image. Specifically, the projection overlap rate between the target obstacle and the vehicle may be calculated based on a formula C1=S3 / S0, wherein S3 represents the area of ​​the target obstacle in the target image, and S0 represents the area of ​​the target image.

[0100] The behavior intention parameter of the target obstacle represents the influence of the behavior intention of the target obstacle on the collision. The larger the behavior intention parameter is, the higher the risk of collision between the vehicle and the target obstacle is. Optionally, the behavior intention parameter of the target obstacle can be obtained according to the pre-travel trajectory of the target obstacle. The pre-travel trajectory of the target obstacle can be detected by the vehicle system. Different pre-travel trajectories correspond to different behavior intention parameters, as shown in Table 4 below:

[0101] Table 4

[0102]

[0103]

[0104] In the above Table 4, if the target obstacle is a motorcycle, when the vehicle system determines that the pre-travel trajectory of the target obstacle is braking, then by looking up the table, it can be determined that the behavior intention parameter is 9, which means that the risk of collision between the vehicle and the target obstacle is high. When the vehicle system determines that the pre-travel trajectory of the target obstacle is turning, then by looking up the table, it can be determined that the behavior intention parameter is 3, which means that the risk of collision between the vehicle and the target obstacle is low. It should be noted that the pre-travel trajectory of the target obstacle and the corresponding behavior intention parameter can be adjusted according to the actual situation, and the above Table 4 is only used as an example.

[0105] In a possible design, the collision risk parameter includes but is not limited to at least one of an obstacle type parameter of the target obstacle, a behavior intention parameter of the target obstacle, a throttle pedal parameter of the vehicle, a projection overlap rate between the target obstacle and the vehicle, a relative distance between the vehicle and the target obstacle, and a relative speed between the vehicle and the target obstacle. In step S202, the actual steering torque of the driver at the current moment and the collision risk parameter can be input into a trained reverse rotation torque prediction model to obtain a first reverse rotation torque for completely offsetting the actual steering torque of the driver at the next moment.

[0106] In a first embodiment, the first reverse rotation torque and the second reverse rotation torque can both be predicted by the reverse rotation torque prediction model, that is, the reverse rotation torque can be predicted in real time by a model, and the reverse rotation torque predicted by the model meets the following requirements:

[0107] When the collision risk value between the vehicle and the target obstacle is greater than or equal to a preset collision risk threshold, the predicted first reverse rotation torque is greater than or equal to the motor steering output torque corresponding to the actual steering operation torque of the driver at the next moment;

[0108] When the collision risk value between the vehicle and the target obstacle is less than a preset collision risk threshold, the predicted second reverse rotation torque is less than the motor steering output torque corresponding to the actual steering operation torque of the driver at the next moment.

[0109] In this embodiment, the collision risk value between the vehicle and the target obstacle can be calculated by the model itself, that is, the model can output the collision risk value between the vehicle and the target obstacle, and at the same time output the predicted reverse rotation torque, as long as the collision risk value and the predicted reverse rotation torque meet the above conditions.

[0110] In a second embodiment, a first reverse rotation torque prediction model and a second reverse rotation torque prediction model obtained through training may be pre-set on the vehicle. When it is determined that the collision risk value between the vehicle and the target obstacle is greater than or equal to a preset collision risk threshold, the actual steering operation torque of the driver at the current moment and the collision risk parameter may be input into the first reverse rotation torque prediction model to obtain a first reverse rotation torque that can completely offset the actual steering operation torque of the driver at the next moment. When it is determined that the collision risk value between the vehicle and the target obstacle is less than the preset collision risk threshold, the actual steering operation torque of the driver at the current moment and the collision risk parameter may be input into the second reverse rotation torque prediction model to obtain a second reverse rotation torque that cannot completely offset the actual steering operation torque of the driver at the next moment.

[0111] The method for calculating the collision risk value in this implementation manner may refer to the above-mentioned method of calculating the collision risk value using the collision risk parameter and each of the weight coefficients, which will not be described in detail here.

[0112] It should be noted that the reverse rotation torque predicted by the reverse rotation torque prediction model in the embodiment of the present application may be positively correlated with the collision risk value between the vehicle and the target obstacle, that is, as the collision risk between the vehicle and the target obstacle increases, the corresponding predicted reverse rotation torque increases. Specifically, the reverse rotation torque predicted by each of the reverse rotation torque prediction models mentioned above is positively correlated with the corresponding collision risk value.

[0113] It is understandable that before the actual steering operation torque of the driver at the current moment and the collision risk parameter are input into the trained reverse rotation torque prediction model, the reverse rotation torque prediction model needs to be trained. The specific training steps may include:

[0114] Step 1: Obtain a training data set and an initial model; the training data set includes the driver's actual steering operation torque of the vehicle at multiple moments, the collision risk parameter between the vehicle and the target obstacle at each moment, and the driver's actual steering operation torque of the vehicle at the next moment after each moment.

[0115] Step 2: Perform training based on the training data set and the initial model.

[0116] Step 3: When the preset training stop condition is reached, the reverse rotation torque prediction model is obtained.

[0117] Among them, the collision risk parameters between the vehicle and the target obstacle at each of the moments include but are not limited to at least one of an obstacle type parameter of the target obstacle, a behavior intention parameter of the target obstacle, an accelerator pedal parameter of the vehicle, a projection overlap rate between the target obstacle and the vehicle, a relative distance between the vehicle and the target obstacle, and a relative speed between the vehicle and the target obstacle.

[0118] After executing step 1 to obtain the training data set, the data in the obtained training data set can be preprocessed to convert each data into a labeled form suitable for model input, and the training data set can be divided into a training set and a validation set.

[0119] In the embodiment of the present application, an initial model can be established, and the established initial model can be a Transformer neural network model. The model includes an encoder and a decoder, the encoder is responsible for understanding the input, and the decoder is responsible for generating the output, wherein the encoder converts the input data into a high-dimensional abstract representation information that can be understood by a computer, and the decoder uses the high-dimensional abstract representation information output by the encoder to generate a target sequence. The weight matrix corresponding to each collision risk parameter of the Transformer model is initialized. For example, the weight matrices corresponding to the obstacle type parameters, the behavior intention parameters, the accelerator pedal parameters, the projection overlap rate, the relative distance, and the relative speed are initialized using a random initialization method using a normal distribution or a uniform distribution.

[0120] Forward propagation: Input the data in the training set into the model. Specifically, input the actual steering torque of the driver at time t and the collision risk parameter between the vehicle and the target obstacle at time t into the Transformer model. According to the architecture of the model, the output of the model is obtained from the input layer through the multi-head attention layer, position encoding, encoding layer, decoding layer, etc., that is, the reverse rotation torque at time t+1 is obtained. The reverse rotation torque at time t+1 is used as the predicted value, and the motor steering output torque corresponding to the actual steering torque of the driver at time t+1 is used as the label value. The loss value is calculated based on the predicted value and the label value. Backward propagation: Use the calculated loss value to calculate the gradient of each weight matrix parameter by the chain rule (reversely traverse each layer of the model to calculate the contribution of each weight matrix parameter to the loss). Weight matrix parameter update: The Adam optimization algorithm is used to update the weight matrix parameters of the model.

[0121] The preset training stop condition in step 3 can be flexibly set by the developer. For example, when the number of training iterations reaches a preset training number threshold, the training is stopped, or when the loss value of the target loss function is lower than a preset loss threshold, the training is stopped.

[0122] Optionally, after determining the reverse rotation torque prediction model, the validation set can be used to evaluate the reverse rotation torque prediction model. When the evaluation is unqualified, the model parameters in the reverse rotation torque prediction model can be readjusted, such as adjusting the model learning rate, the number of neural network layers, etc., and re-training can be performed until a qualified reverse rotation torque prediction model is obtained.

[0123] Optionally, the validation set includes the actual steering torque of the driver of the vehicle at multiple moments, the collision risk parameter between the vehicle and the target obstacle at each moment, and the actual steering torque of the driver of the vehicle at the next moment of each moment. It is understandable that the specific time corresponding to the moment in the validation set is different from the specific time corresponding to the moment in the training set. Exemplarily, the verification set includes the actual steering operation torque of the driver of the vehicle at time i, the collision risk parameter between the vehicle and the target obstacle at time i, and the actual steering operation torque of the driver of the vehicle at time i+1; the embodiment of the present application can calculate the collision risk value of the vehicle and the target obstacle based on the various collision risk parameters in the verification set. If the collision risk value is within a preset high-risk numerical range, for example, the collision risk value is greater than or equal to a preset collision risk threshold, a first reverse rotation torque is calculated based on the reverse rotation torque prediction model, and it is determined whether the difference between the first reverse rotation torque and the motor steering output torque corresponding to the actual steering operation torque of the driver at time i+1 exceeds the first preset difference threshold. If so, it indicates that the first reverse rotation torque is much greater than the motor steering output torque, and the model meets the first qualification condition. If not, the reverse rotation torque prediction model is retrained according to the above-described process. If the collision risk value is within the preset low-risk numerical range, for example, the collision risk value is less than the preset collision risk threshold, a second reverse rotation torque is calculated based on the reverse rotation torque prediction model, and it is determined whether the difference between the second reverse rotation torque and the motor steering output torque corresponding to the actual steering operation torque of the driver at time i+1 is lower than the second preset difference threshold. If so, it indicates that the second reverse rotation torque is much lower than the motor steering output torque, and the model meets the second qualification condition. If not, the reverse rotation torque prediction model is retrained according to the above-described process. Exemplarily, when the model is evaluated through a validation set, if the proportion of data in the validation set that meets the first qualification condition or the second qualification condition reaches the preset proportion threshold, the model is determined to be qualified. In this way, it can also be verified whether the reverse rotation torque predicted by the reverse rotation torque prediction model is positively correlated with the collision risk between the vehicle and the target obstacle.

[0124] The flowchart of training the prediction model in the embodiment of the present application is as follows: Figure 4 As shown, in Figure 4 In the process, the actual steering operation torque of the driver and various collision risk parameters are input into the initial model for training, the reverse rotation torque is predicted, and the model output results are evaluated. The model output results include: reverse rotation torque and loss value. The loss value can react to the model training so that the prediction model can be optimized based on the loss value. When the loss value is lower than the preset loss threshold, the final reverse rotation torque prediction model is determined.

[0125] In one embodiment, a logic flow chart for preventing a vehicle collision is provided, referring to Figure 5 At time t, the actual steering operation torque TQ0 of the driver and the collision risk parameters are obtained. The collision risk parameters include: obstacle type parameters of the target obstacle, relative distance, relative speed, projection overlap rate between the target obstacle and the vehicle, behavior intention parameters of the target obstacle and accelerator pedal parameters. The reverse rotation torque TQ1 at time t+1 is predicted based on the various collision risk parameters and TQ0. At time t+1, the magnitude of TQ1 and the motor steering output torque TQ3 is judged. TQ3 is obtained based on TQ0 at time t+1. When TQ1≥TQ3, the driver cannot operate the steering wheel to steer the vehicle; when TQ1<TQ3, the driver can operate the steering wheel to steer, thereby being able to assist the vehicle in safe driving in real time based on the reverse rotation torque, thereby ensuring the safety of the vehicle during driving.

[0126] It is understandable that in the embodiment of the present application, after the AES function is activated, automatic emergency steering can be achieved through the AES function. After the AES function of the vehicle is activated, the third reverse rotation torque used to completely offset the driver's actual steering operation torque at the next moment can continue to be predicted based on the driver's actual steering operation torque at the current moment (latest moment) and the collision risk parameter. When the next moment arrives, the driver's actual steering operation torque at the next moment is completely offset by the third reverse rotation torque. This can prevent the AES function from exiting due to abnormality due to the driver's misoperation of the steering wheel after the AES function is activated, so as to better ensure that obstacle avoidance can be achieved through the AES function.

[0127] In one embodiment, the present application embodiment provides a schematic diagram of an AES scenario operation, referring to Figure 6 , the reverse rotation torque prediction model is used to predict the collision risk value of the vehicle and the reverse rotation torque at the next moment in real time. The vehicle is determined to have no collision risk in state one, low collision risk in state two, and high collision risk in state three. At this time, the vehicle automatically turns on the AES function and implements automatic emergency steering. If the driver operates the steering wheel during the automatic driving of the vehicle, the steering of the vehicle will be controlled based on the reverse rotation torque and the motor steering output torque of the vehicle motor, so that the driver cannot operate the steering wheel to implement the vehicle steering, thereby ensuring the successful activation of the AES function and preventing the vehicle from colliding. When the vehicle successfully changes lanes, state four is no collision risk, the AES function is turned off, the driver can operate the steering wheel, and the vehicle is transformed from automatic driving to manual driving.

[0128] Based on the above method, when the collision risk value between the vehicle and the target obstacle is greater than or equal to the preset collision risk threshold, the first reverse rotation torque at the next moment is predicted according to the actual steering operation torque of the driver at the current moment of the vehicle and the collision risk parameter. When the next moment arrives, the actual steering operation torque of the driver at the next moment is completely offset by the first reverse rotation torque, and the AES function of the vehicle is activated. Through the above method, when the next moment arrives, since the first reverse rotation torque can completely offset the actual steering operation torque of the driver at the next moment, the problem of the AES function being unable to be activated and causing a collision accident due to the driver's manipulation of the steering wheel when the next moment arrives can be solved, which effectively prevents the vehicle from colliding and improves the safety of the vehicle during driving.

[0129] It should be understood that although Figure 2 The steps in the flowchart are shown in sequence as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified in this document, there is no strict order restriction for the execution of these steps, and these steps can be executed in other orders. Moreover, Figure 2 At least part of the steps may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily executed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least part of the sub-steps or stages of other steps.

[0130] In one embodiment, Figure 7 As shown, a device for preventing vehicle collision is provided, comprising: an acquisition module 701, a prediction module 702 and a control module 703, wherein:

[0131] An acquisition module 701 is used to acquire the actual steering operation torque of the driver of the vehicle at the current moment, and a collision risk parameter between the vehicle and the target obstacle at the current moment;

[0132] A prediction module 702 is configured to predict, when the collision risk value between the vehicle and the target obstacle is greater than or equal to a preset collision risk threshold, a first reverse rotation torque for completely offsetting the actual steering operation torque of the driver at a next moment according to the actual steering operation torque of the driver at the current moment and the collision risk parameter;

[0133] The control module 703 is used for completely offsetting the actual steering operation torque of the driver at the next moment by the first reverse rotation torque when the next moment arrives, and activating the AES function of the vehicle.

[0134] In one embodiment, the control module 703 is used to convert the actual steering operation torque of the driver at the next moment to obtain the corresponding motor steering output torque when the next moment arrives; the first reverse rotation torque is greater than or equal to the motor steering output torque; when the first reverse rotation torque is greater than the motor steering output torque, the vehicle is controlled to turn in the direction of the first reverse rotation torque based on the difference between the first reverse rotation torque and the motor steering output torque; when the first reverse rotation torque is equal to the motor steering output torque, the driving direction of the vehicle is maintained consistent with the driving direction at the previous moment.

[0135] In one embodiment, the collision risk parameters include at least one of an obstacle type parameter of the target obstacle, a behavior intention parameter of the target obstacle, an accelerator pedal parameter of the vehicle, a projection overlap rate between the target obstacle and the vehicle, a relative distance between the vehicle and the target obstacle, and a relative speed between the vehicle and the target obstacle; the prediction module 702 is used to input the actual steering operation torque of the driver at the current moment and the collision risk parameters into a trained reverse rotation torque prediction model to obtain a first reverse rotation torque for completely offsetting the actual steering operation torque of the driver at the next moment.

[0136] In one embodiment, the device includes a training module for acquiring a training data set and an initial model; the training data set includes the actual steering operation torque of the driver of the vehicle at multiple moments, the collision risk parameters between the vehicle and the target obstacle at each of the moments, and the actual steering operation torque of the driver of the vehicle at the next moment of each of the moments; training is performed based on the training data set and the initial model; when a preset training stop condition is reached, the reverse rotation torque prediction model is obtained.

[0137] In one embodiment, the reverse rotation torque predicted by the reverse rotation torque prediction model is positively correlated with a collision risk value between the vehicle and the target obstacle.

[0138] In one embodiment, the prediction module 702 is also used to predict a second reverse rotation torque by using the reverse rotation torque prediction model, the actual steering operation torque of the driver at the current moment and the collision risk parameter when the collision risk value between the vehicle and the target obstacle is less than the preset collision risk threshold; the control module 703 is used to convert the actual steering operation torque of the driver at the next moment to obtain the corresponding motor steering output torque when the next moment arrives; the second reverse rotation torque is less than the motor steering output torque; and the vehicle is controlled to steer in the direction of the motor steering output torque according to the difference between the motor steering output torque and the second reverse rotation torque.

[0139] In one embodiment, the collision risk parameters include at least two of an obstacle type parameter of the target obstacle, a behavior intention parameter of the target obstacle, an accelerator pedal parameter of the vehicle, a projection overlap rate between the target obstacle and the vehicle, a relative distance between the vehicle and the target obstacle, and a relative speed between the vehicle and the target obstacle; the prediction module 702 is used to obtain a weight coefficient corresponding to each of the collision risk parameters; and based on each of the collision risk parameters and each of the weight coefficients, calculate a collision risk value between the vehicle and the target obstacle.

[0140] The specific definition of the device for preventing vehicle collision can be found in the definition of the method for preventing vehicle collision mentioned above, which will not be repeated here. Each module in the above-mentioned device for preventing vehicle collision can be implemented in whole or in part by software, hardware and a combination thereof. The above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory in the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0141] In one embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as follows: Figure 8 As shown. The computer device includes a processor, a memory, a network interface and a database connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store data for preventing vehicle collisions. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a method for preventing vehicle collisions is implemented.

[0142] Those skilled in the art will understand that Figure 8 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0143] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps in any of the above-described methods when executing the computer program.

[0144] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps in any of the above-mentioned methods are implemented.

[0145] The processor can be an integrated circuit chip with signal processing capabilities. The above processor can be a general-purpose processor, including a CPU (Central Processing Unit), NP (Network Processor), etc.; it can also be a DSP (Digital Signal Processor), ASIC (Application Specific Integrated Circuit), FPGA (Field Programmable Gate Array) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. It can implement or execute the methods, steps and logic block diagrams disclosed in the embodiments of the present application. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc.

[0146] The memory may include but is not limited to RAM (Random Access Memory), ROM (Read Only Memory), PROM (Programmable Read Only Memory), EPROM (Erasable Programmable Read-Only Memory) and EEPROM (Electrically Erasable Programmable Read Only Memory).

[0147] It should be noted that the diagram provided in the present embodiment only illustrates the basic concept of the present invention in a schematic manner, so the diagram only shows the components related to the present invention rather than drawing according to the number, shape and size of the components during actual implementation. The type, quantity and ratio of each component during actual implementation can be a random change, and the component layout type may also be more complicated. The structure, ratio, size, etc. illustrated in the drawings of the present specification are only used to match the content disclosed in the specification for people familiar with this technology to understand and read, and are not used to limit the limiting conditions that the present invention can implement, so they have no technical substantive significance. Any modification of the structure, change of the proportional relationship or adjustment of the size should still fall within the scope of the technical content disclosed by the present invention without affecting the effect that the present invention can produce and the purpose that can be achieved. At the same time, the terms such as "upper", "lower", "left", "right", "middle" and "one" quoted in this specification are only for the convenience of narration, and are not used to limit the scope of the present invention. The change or adjustment of its relative relationship should also be regarded as the scope of the present invention without substantially changing the technical content.

[0148] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0149] The above-mentioned embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be subject to the attached claims.

Claims

1. A method for preventing vehicle collision, characterized in that: include: Acquiring the actual steering operation torque of the driver of the vehicle at the current moment, and the collision risk parameter between the vehicle and the target obstacle at the current moment; When the collision risk value between the vehicle and the target obstacle is greater than or equal to a preset collision risk threshold, predicting a first reverse rotation torque for completely offsetting the actual steering operation torque of the driver at a next moment according to the actual steering operation torque of the driver at the current moment and the collision risk parameter; When the next moment arrives, the driver's actual steering operation torque at the next moment is completely offset by the first reverse rotation torque, and the AES function of the vehicle is activated.

2. The method for preventing vehicle collision according to claim 1, characterized in that: When the next moment arrives, the first reverse rotation torque completely offsets the actual steering operation torque of the driver at the next moment, including: When the next moment arrives, the actual steering operation torque of the driver at the next moment is converted to obtain the corresponding motor steering output torque; the first reverse rotation torque is greater than or equal to the motor steering output torque; When the first reverse rotation torque is greater than the motor steering output torque, controlling the vehicle to steer in the direction of the first reverse rotation torque based on a difference between the first reverse rotation torque and the motor steering output torque; When the first reverse rotation torque is equal to the motor steering output torque, the driving direction of the vehicle is maintained consistent with the driving direction at the last moment.

3. The method for preventing vehicle collision according to claim 1, characterized in that: The collision risk parameter includes at least one of an obstacle type parameter of the target obstacle, a behavior intention parameter of the target obstacle, an accelerator pedal parameter of the vehicle, a projection overlap ratio between the target obstacle and the vehicle, a relative distance between the vehicle and the target obstacle, and a relative speed between the vehicle and the target obstacle; The predicting, based on the actual steering operation torque of the driver at the current moment and the collision risk parameter, a first reverse rotation torque for completely offsetting the actual steering operation torque of the driver at the next moment comprises: The actual steering operation torque of the driver at the current moment and the collision risk parameter are input into the trained reverse rotation torque prediction model to obtain the first reverse rotation torque.

4. The method for preventing vehicle collision according to claim 3, characterized in that: Before inputting the actual steering operation torque of the driver at the current moment and the collision risk parameter into the trained reverse rotation torque prediction model, the method includes: Acquire a training data set and an initial model; the training data set includes the actual steering operation torque of the driver of the vehicle at multiple moments, the collision risk parameter between the vehicle and the target obstacle at each moment, and the actual steering operation torque of the driver of the vehicle at the next moment of each moment; Performing training based on the training data set and the initial model; When the preset training stop condition is reached, the reverse rotation torque prediction model is obtained.

5. The method for preventing vehicle collision according to claim 3, characterized in that: The reverse rotation torque predicted by the reverse rotation torque prediction model is positively correlated with a collision risk value between the vehicle and the target obstacle.

6. The method for preventing vehicle collision according to claim 3, characterized in that: The method further comprises: When the collision risk value between the vehicle and the target obstacle is less than the preset collision risk threshold, the reverse rotation torque prediction model predicts a second reverse rotation torque according to the actual steering operation torque of the driver at the current moment and the collision risk parameter; When the next moment arrives, the actual steering operation torque of the driver at the next moment is converted to obtain the corresponding motor steering output torque; the second reverse rotation torque is less than the motor steering output torque; The vehicle is controlled to steer in the direction of the motor steering output torque according to a difference between the motor steering output torque and the second reverse rotation torque.

7. The method for preventing vehicle collision according to claim 1, characterized in that: The collision risk parameter includes at least two of an obstacle type parameter of the target obstacle, a behavior intention parameter of the target obstacle, an accelerator pedal parameter of the vehicle, a projection overlap rate between the target obstacle and the vehicle, a relative distance between the vehicle and the target obstacle, and a relative speed between the vehicle and the target obstacle; the collision risk value is obtained by: Obtaining a weight coefficient corresponding to each of the collision risk parameters; Based on the collision risk parameters and the weight coefficients, a collision risk value between the vehicle and the target obstacle is calculated.

8. A device for preventing vehicle collision, characterized in that: The device comprises: An acquisition module, used to acquire the actual steering operation torque of the driver of the vehicle at the current moment, and a collision risk parameter between the vehicle and the target obstacle at the current moment; a prediction module, configured to predict, when a collision risk value between the vehicle and the target obstacle is greater than or equal to a preset collision risk threshold, a first reverse rotation torque for completely offsetting the actual steering operation torque of the driver at a next moment according to the actual steering operation torque of the driver at the current moment and the collision risk parameter; The control module is used for completely offsetting the actual steering operation torque of the driver at the next moment by the first reverse rotation torque when the next moment arrives, and activating the AES function of the vehicle.

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

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

Images