Reaction model construction method, parameter calibration method for reaction model and related apparatus

By introducing the concept of potential risk points, building a response model and covering more dangerous driving scenarios, the problem that the existing technology cannot effectively evaluate the autonomous driving capabilities and improve the safety and evaluation accuracy of autonomous driving.

WO2024235008A1PCT designated stage expired Publication Date: 2024-11-21YINWANG INTELLIGENT TECHNOLOGIES CO LTD

Patent Information

Application Number
PCT/CN2024/090697
Authority / Receiving Office
WO · WO
Patent Type
Applications
Current Assignee / Owner
Priority Date
2023-05-15
Filing Date
2024-04-29
Publication Date
2024-11-21

AI Technical Summary

Technical Problem

The existing autonomous driving response model cannot effectively cover most dangerous driving scenarios, resulting in inaccurate assessment of autonomous driving capabilities and affecting safety.

Method used

Introduce the concept of potential risk points, build a reaction model by obtaining information on real and potential risk points, cover more dangerous driving scenarios, optimize action behavior logic, comprehensive braking and steering behavior, and improve the accuracy of evaluation.

Benefits of technology

It realizes a more reasonable assessment of autonomous driving capabilities, improves the safety of autonomous driving, and can more accurately characterize the vehicle's movement behavior after a risk point occurs.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN2024090697_21112024_PF_FP_ABST
    Figure CN2024090697_21112024_PF_FP_ABST
Patent Text Reader

Abstract

A reaction model construction method, a parameter calibration method for a reaction model and a related apparatus, which are applied to the technical field of intelligent driving. The method comprises: acquiring first information, and constructing a reaction model on the basis of the first information. The first information comprises information of a first risk point or a second risk point, the first risk point being a real risk point appearing on a road where a first vehicle is located at a current moment, the second risk point being a potential risk point appearing on the road where the first vehicle is located at the current moment which triggers a third risk point, and the third risk point being a real risk point appearing on the road where the first vehicle is located within a first time period after the current moment; the reaction model is used to represent action behaviors of the first vehicle after the first risk point or the second risk point appears. The present method can cover most dangerous driving scenarios, thereby reasonably evaluating automatic driving capability, and improving automatic driving safety.
Need to check novelty before this filing date? Find Prior Art

Description

Reaction model construction method, reaction model parameter calibration method and related device

[0001] This application claims priority to the Chinese patent application filed with the State Intellectual Property Office of China on May 15, 2023, with application number 202310546008.7 and application name “Reaction model construction method, reaction model parameter calibration method and related devices”, the entire contents of which are incorporated by reference into this application. Technical Field

[0002] The present application relates to the field of intelligent driving technology, and in particular to a reaction model construction method, a reaction model parameter calibration method, and related devices. Background Art

[0003] Autonomous driving technology is based on artificial intelligence (AI). It uses sensors, cameras, radar, lasers, and other sensing devices to perceive and identify the environment, enabling autonomous driving and significantly improving driving efficiency. With the rapid development of autonomous driving technology, its safety is drawing increasing attention.

[0004] The current industry consensus on autonomous driving capabilities is that they must avoid accidents that are reasonably foreseeable and avoidable within their designed operating range. A common reference model for evaluating "avoidability" is the reaction model of a careful and competent human driver (CC Driver). This reaction model can be used to assess whether a potential accident in a driving scenario is avoidable for a typical human driver, thereby evaluating the autonomous driving capability.

[0005] However, the current response model still cannot cover most dangerous driving scenarios, resulting in unreasonable assessments of autonomous driving capabilities and affecting the safety of autonomous driving.

[0006] Summary of the Invention

[0007] The embodiments of the present application provide a reaction model construction method, a reaction model parameter calibration method and related devices, which can cover most dangerous driving scenarios, thereby making a reasonable assessment of the autonomous driving capability and improving the safety of autonomous driving.

[0008] In a first aspect, an embodiment of the present application provides a reaction model construction method, which is applied to a first vehicle and can be performed by a communication device. The communication device can be a device, or a chip (system) or circuit for a device, which is not limited in this application. The method includes:

[0009] Obtaining first information, the first information including information about a first risk point or a second risk point, the first risk point being a real risk point occurring on the road where the first vehicle is located at a current moment, the second risk point being a potential risk point occurring on the road where the first vehicle is located at a current moment that may trigger a third risk point, and the third risk point being a real risk point occurring on the road where the first vehicle is located within a first time period after the current moment;

[0010] A reaction model is constructed based on the first information, where the reaction model is used to characterize the action behavior of the first vehicle after the first risk point or the second risk point occurs.

[0011] In an embodiment of the present application, a reaction model construction method is provided, where a communication device obtains first information and constructs a reaction model based on the first information. The communication device here can also be a processor / chip that can be used to execute computer execution instructions, and the embodiment of the present application does not limit this.

[0012] The first information in the embodiment of the present application includes information about the first risk point or the second risk point. Among them, the first risk point is a real risk point that appears on the road where the first vehicle is located at the current moment. The real risk point can be understood as an event that threatens the driving safety of the vehicle at the current moment. The second risk point is a potential risk point that appears on the road where the first vehicle is located at the current moment and triggers the third risk point. The third risk point is a real risk point that appears on the road where the first vehicle is located in the first time period after the current moment. The potential risk point can be understood as an event that may affect the driving safety of the vehicle but has not yet posed an actual threat to the driving safety of the vehicle. The subsequent events triggered by it pose a threat to the driving safety of the vehicle. It can be understood that the second risk point can be used to provide a warning for a reasonable and foreseeable real risk point in the first time period after the current moment. The reaction model constructed based on the above-mentioned first information in the embodiment of the present application is used to characterize the action behavior of the first vehicle after the above-mentioned first risk point or the above-mentioned second risk point appears.

[0013] In the embodiment of the present application, a method is designed to quickly construct a reaction model of a general human driver with concentrated attention in different autonomous driving scenarios. By introducing the concept of potential risk points, more dangerous driving scenarios can be covered, and the definition logic of real risk points can be optimized to more accurately characterize the vehicle's action behavior after the risk points appear, thereby making a reasonable assessment of the autonomous driving capabilities and improving the safety of autonomous driving.

[0014] In combination with the above first aspect, in one possible implementation, the second risk point includes any one or more of the following:

[0015] Road sections with deceleration warnings, road sections with attention signs, road sections with static blind spots, vehicles in adjacent lanes braking, large adjacent vehicles, and warning signals of adjacent vehicles.

[0016] In an embodiment of the present application, a possible specific implementation of a second risk point is provided. Specifically, the second risk point may include, but is not limited to, a section of the road where the first vehicle is located with a deceleration warning, a section of the road with a caution sign, a section of the road with a static blind spot, a vehicle in an adjacent lane braking, a large adjacent vehicle, a warning signal from an adjacent vehicle, and so on. It is understandable that the above-mentioned potential risk situations can be used to provide a foreshadowing warning of a real risk point that is reasonably foreseeable within a first time period after the current moment. By introducing the concept of potential risk points, the embodiments of the present application can cover more dangerous driving scenarios, thereby making a reasonable assessment of the capabilities of autonomous driving and improving the safety of autonomous driving.

[0017] In combination with the above-mentioned first aspect, in a possible implementation manner, the action behavior includes at least one of a vehicle braking behavior or a vehicle steering behavior.

[0018] In an embodiment of the present application, a possible specific implementation of the first vehicle's behavior after the occurrence of a first risk point or a second risk point is provided. Specifically, the behavior may include, but is not limited to, at least one of vehicle braking or vehicle steering. The reaction model of a typical human driver designed in the embodiment of the present application integrates the two action logics of steering and braking, comprehensively and realistically reflecting the differences in the defensive and stress responses of a typical human driver, and more accurately characterizing the vehicle's behavior after the occurrence of a risk point, thereby making a reasonable assessment of the autonomous driving capability and improving the safety of autonomous driving.

[0019] In combination with the first aspect above, in a possible implementation, the first risk point is that the offset between the second vehicle and the center line of the lane where the second vehicle is located is greater than a first threshold; wherein, the second vehicle is a vehicle that cuts in front of the first vehicle without turning on the turn signal, and the model of the second vehicle is not larger than the model of the first vehicle.

[0020] In an embodiment of the present application, a possible specific implementation of the first risk point is provided, specifically, in a scenario where a second vehicle cuts in front of a first vehicle without turning on the turn signal, and the model of the second vehicle is not larger than that of the first vehicle, if the offset between the second vehicle and the center line of the lane where the second vehicle is located is greater than a first threshold, it is determined to be a first risk point. The first threshold is not a fixed value and can be adjusted according to different application scenarios. It can be understood that in the scenario where the second vehicle cuts in front of the first vehicle without turning on the turn signal in the embodiment of the present application, the second vehicle does not turn on the turn signal, is not a large vehicle, and does not perform braking actions, so it can be considered that there are no potential risk points in this scenario, that is, it does not pose a potential risk to the driving safety of the first vehicle. The first risk point at this time can be the moment when the second vehicle shows an obvious intention to cut in. Accordingly, the reaction model constructed based on the first information is used to characterize the defenseless emergency action behavior of the first vehicle in this scenario after the first risk point appears.

[0021] In combination with the above-mentioned first aspect, in a possible implementation manner, the second risk point is that the model of the third vehicle is larger than that of the first vehicle, and the third risk point is that the third vehicle is laterally offset; wherein, the third vehicle is adjacent to the first vehicle, and the third vehicle is a vehicle that cuts in front of the first vehicle without turning on the turn signal.

[0022] In an embodiment of the present application, a possible specific implementation of a second risk point is provided. Specifically, in a scenario where a third vehicle cuts in front of a first vehicle without turning on its turn signal, if the third vehicle is larger than the first vehicle, the third vehicle can be identified as a second risk point before the third vehicle cuts in. If the third vehicle deviates laterally, it can be identified as a third risk point. It is understood that in the scenario where the third vehicle cuts in front of the first vehicle without turning on its turn signal in this embodiment of the present application, although the third vehicle does not turn on its turn signal, it is a large vehicle adjacent to the first vehicle. Therefore, this scenario can be considered to present a potential risk point, i.e., a potential risk to the driving safety of the first vehicle. In this case, the second risk point can be identified as the large vehicle adjacent to the first vehicle (i.e., the third vehicle). Since the third vehicle is a large vehicle, its cutting in action is more obvious. Therefore, the lateral deviation of the third vehicle can be identified as the third risk point, i.e., the actual risk point that may be triggered by the second risk point. It is understood that the second risk point (the large vehicle adjacent to the first vehicle) can be used to provide a forewarning of a reasonably foreseeable actual risk point (the lateral deviation of the third vehicle) within a first time period after the current moment. Accordingly, the reaction model constructed based on the first information is used to characterize the defensive emergency action behavior of the first vehicle in the scenario after the second risk point appears.

[0023] In combination with the above-mentioned first aspect, in a possible implementation manner, the first risk point is that the lateral body of the fourth vehicle appearing within the field of view of the first vehicle is greater than a second threshold; wherein, the fourth vehicle, the first vehicle and the fifth vehicle are located in the same lane, and the fourth vehicle is located in front of the fifth vehicle, the fifth vehicle is located in front of the first vehicle, and the fifth vehicle is the vehicle that cuts out of the current lane.

[0024] In an embodiment of the present application, a possible specific implementation of a first risk point is provided, specifically, when the fifth vehicle in front of the first vehicle cuts out of the current lane, in the scenario where the first vehicle encounters the fourth vehicle in front of the fifth vehicle, if the lateral body of the fourth vehicle that appears within the field of view of the first vehicle is greater than the second threshold, it is determined as the first risk point. The second threshold is not a fixed value and can be adjusted according to different application scenarios. It can be understood that in the scenario where the fifth vehicle in the embodiment of the present application cuts out of the current lane and the first vehicle encounters the fourth vehicle in front of the fifth vehicle, there is no potential risk point, that is, it does not pose a potential risk to the driving safety of the first vehicle. The first risk point at this time can be the moment when the lateral body of the fourth vehicle that appears within the field of view of the first vehicle is greater than the second threshold. Accordingly, the reaction model constructed based on the first information is used to characterize the defenseless emergency action behavior of the first vehicle in this scenario after the first risk point appears.

[0025] In conjunction with the first aspect above, in one possible implementation, constructing a reaction model based on the first information includes:

[0026] When the first information includes the first risk point, second information is determined, where the second information includes information about a first interval and a second interval, the first interval being an interval in which the first vehicle is traveling at a constant speed, the second interval being an interval in which the first vehicle is braking, and the second interval being located after the first interval;

[0027] Based on the first information and the second information, a first reaction model is constructed. The first reaction model is used to characterize that after the first risk point occurs, the first vehicle travels at a constant speed in the first interval and brakes until it stops in the second interval.

[0028] In an embodiment of the present application, a possible specific implementation method for constructing a first reaction model is provided. Specifically, when the first information includes a first risk point, second information is determined, and the first reaction model is constructed based on the first and second information. When the first information includes the first risk point, it indicates that a real risk point has appeared on the road where the first vehicle is currently located. In this case, the determined second information includes information about the first interval in which the first vehicle is traveling at a constant speed and information about the second interval in which the vehicle is braking. Accordingly, the first reaction model constructed based on the first and second information is used to characterize the first vehicle's defenseless emergency braking behavior after the first risk point appears in this scenario. That is, after the first risk point appears, the first vehicle continues to travel at a constant speed in the first interval, then begins braking in the second interval, reaches the maximum braking deceleration at a certain slope, and maintains deceleration until the vehicle stops. The first reaction model constructed by the embodiment of the present application can cover driving scenarios where real risk points appear, optimize the definition logic of real risk points, and more accurately characterize the vehicle's defenseless emergency braking behavior after the real risk point appears, thereby making a reasonable assessment of the capability of autonomous driving and improving the safety of autonomous driving.

[0029] In combination with the foregoing first aspect, in a possible implementation manner, the second information further includes any one or more of the following:

[0030] The duration of the constant speed travel of the first vehicle in the first interval, the start time point of vehicle braking of the first vehicle in the second interval, the braking rate of the first vehicle, and the maximum braking deceleration of the first vehicle.

[0031] In an embodiment of the present application, a possible specific implementation of the second information is provided. Specifically, when the first information includes a first risk point, the second information required to construct the first reaction model may include, but is not limited to, the duration of the first vehicle's constant speed travel within the first interval, the start time of vehicle braking by the first vehicle within the second interval, the braking rate of the first vehicle, the maximum braking deceleration of the first vehicle, and so on. The first reaction model constructed based on the second information in the embodiment of the present application can more accurately characterize the vehicle's defenseless emergency braking behavior after the occurrence of a real risk point.

[0032] In conjunction with the first aspect above, in one possible implementation, constructing a reaction model based on the first information includes:

[0033] determining, when the first information includes the second risk point, third information, the third information including information about a third interval and a fourth interval, the third interval being an interval in which the first vehicle is traveling at a constant speed, the fourth interval being an interval in which the first vehicle is braking, and the fourth interval being located after the third interval;

[0034] Based on the first information and the third information, a second reaction model is constructed. The second reaction model is used to characterize that after the second risk point occurs and before the third risk point occurs, the first vehicle travels in the current driving state, and after the third risk point occurs, the first vehicle travels at a constant speed in the third interval and brakes until it stops in the fourth interval.

[0035] In an embodiment of the present application, a possible specific implementation for constructing a second reaction model is provided. Specifically, when the first information includes a second risk point, third information is determined, and the second reaction model is constructed based on the first and third information. When the first information includes the second risk point, it indicates that a potential risk point has appeared on the road where the first vehicle is currently located. This can be used to provide a forewarning of a reasonably foreseeable third risk point (i.e., a real risk point) within a first time period after the current moment. In this case, the determined third information includes information about the third interval in which the first vehicle is traveling at a constant speed and information about the fourth interval in which the vehicle is braking. Accordingly, the second reaction model constructed based on the first and third information is used to characterize the defensive emergency braking behavior of the first vehicle after the second risk point appears in this scenario. That is, after the second risk point appears and before the third risk point appears, the first vehicle maintains a coasting state, i.e., driving in its current driving state. After the third risk point appears, the first vehicle, after continuing to travel at a constant speed in the third interval, begins braking in the fourth interval, reaching a maximum braking deceleration at a certain slope, and maintaining the vehicle's deceleration until it stops. The second reaction model constructed through the embodiment of the present application introduces the concept of potential risk points, which can cover driving scenarios where potential risk points appear, and can optimize the definition logic of real risk points, and more accurately characterize the vehicle's defensive emergency braking behavior after the potential risk points appear, thereby making a reasonable assessment of the autonomous driving capability and improving the safety of autonomous driving.

[0036] In combination with the above first aspect, in a possible implementation manner, the third information further includes any one or more of the following:

[0037] The duration of the constant speed travel of the first vehicle in the third interval, the starting time point of vehicle braking of the first vehicle in the fourth interval, the braking rate of the first vehicle, and the maximum braking deceleration of the first vehicle.

[0038] In an embodiment of the present application, a possible specific implementation of the third information is provided. Specifically, when the first information includes a second risk point, the third information required to construct the second reaction model may include, but is not limited to, the duration of the first vehicle's constant speed travel within the third interval, the start time of the first vehicle's vehicle braking within the fourth interval, the first vehicle's braking rate, the first vehicle's maximum braking deceleration, and so on. The second reaction model constructed based on the third information in the embodiment of the present application can more accurately characterize the vehicle's defensive emergency braking behavior after the potential risk point occurs.

[0039] In conjunction with the first aspect above, in one possible implementation, constructing a reaction model based on the first information includes:

[0040] determining fourth information when the first information includes the first risk point, the fourth information including information about a fifth interval and a sixth interval, the fifth interval being an interval in which the first vehicle travels in a straight line, the sixth interval being an interval in which the first vehicle turns, and the sixth interval being located after the fifth interval;

[0041] A third reaction model is constructed based on the first information and the fourth information. The third reaction model is used to characterize that after the first risk point occurs, the first vehicle travels in a straight line within the fifth interval and turns in the sixth interval until returning to the normal direction.

[0042] In an embodiment of the present application, a possible specific implementation for constructing a third reaction model is provided. Specifically, when the first information includes a first risk point, fourth information is determined, and the third reaction model is constructed based on the first and fourth information. When the first information includes the first risk point, it indicates that a real risk point has occurred on the road where the first vehicle is currently located. In this case, the determined fourth information includes information about the fifth interval in which the first vehicle is traveling in a straight line and information about the sixth interval in which the vehicle is turning. Accordingly, the third reaction model constructed based on the first and fourth information is used to characterize the first vehicle's defenseless emergency steering behavior after the first risk point occurs in this scenario. That is, after the first risk point occurs, the first vehicle continues traveling in a straight line in the fifth interval and then begins turning in the sixth interval, sinusoidally reaching a maximum turning angle and then returning to a positive angle, maintaining the direction after returning to a positive angle and continuing to travel in a straight line. The third reaction model constructed by this embodiment of the present application can cover driving scenarios where real risk points occur, optimize the definition logic of real risk points, and more accurately characterize the vehicle's defenseless emergency steering behavior after the occurrence of real risk points, thereby providing a reasonable assessment of the autonomous driving capability and improving the safety of autonomous driving.

[0043] In combination with the foregoing first aspect, in a possible implementation manner, the fourth information further includes any one or more of the following:

[0044] duration of the straight-line driving of the first vehicle in the fifth interval, a peak steering wheel angle of the first vehicle during vehicle steering in the sixth interval, and a steering wheel rotation cycle of the first vehicle.

[0045] In an embodiment of the present application, a possible specific implementation of the fourth information is provided. Specifically, when the first information includes the first risk point, the fourth information required to construct the third reaction model may include, but is not limited to, the duration of the first vehicle's straight-line travel within the fifth interval, the peak steering wheel angle of the first vehicle during vehicle steering within the sixth interval, the steering wheel rotation period of the first vehicle, and so on. The third reaction model constructed based on the fourth information in the embodiment of the present application can more accurately characterize the vehicle's defenseless emergency steering behavior after the occurrence of a real risk point.

[0046] In conjunction with the first aspect above, in one possible implementation, constructing a reaction model based on the first information includes:

[0047] determining fifth information when the first information includes the second risk point, the fifth information including information about a seventh interval and an eighth interval, the seventh interval being an interval in which the first vehicle travels in a straight line, the eighth interval being an interval in which the first vehicle turns, and the eighth interval being located after the seventh interval;

[0048] Based on the first information and the fifth information, a fourth reaction model is constructed. The fourth reaction model is used to characterize that after the second risk point occurs and before the third risk point occurs, the first vehicle travels in the current driving state, and after the third risk point occurs, the first vehicle travels in a straight line within the seventh interval and turns the vehicle within the eighth interval until it returns to the normal direction.

[0049] In an embodiment of the present application, a possible specific implementation for constructing a fourth reaction model is provided. Specifically, when the first information includes a second risk point, fifth information is determined, and the fourth reaction model is constructed based on the first and fifth information. When the first information includes the second risk point, it indicates that a potential risk point has appeared on the road where the first vehicle is currently located. This information can be used to provide a forewarning of a reasonably foreseeable third risk point (i.e., a real risk point) within a first time period after the current moment. In this case, the determined fifth information includes information from the seventh interval in which the first vehicle is traveling in a straight line and information from the eighth interval in which the vehicle is turning. Accordingly, the fourth reaction model constructed based on the first and fifth information is used to characterize the defensive emergency turning behavior of the first vehicle after the second risk point appears in this scenario. Specifically, after the second risk point appears and before the third risk point appears, the first vehicle maintains a coasting state, i.e., driving in its current driving state. After the third risk point appears, the first vehicle, after continuous straight-line driving in the seventh interval, begins turning in the eighth interval, sinusoidally reaching a maximum turning angle and then returning to a positive angle, maintaining the direction after returning to a positive angle and continuing straight-line driving. The fourth reaction model constructed through the embodiment of the present application introduces the concept of potential risk points, which can cover driving scenarios where potential risk points appear, and can optimize the definition logic of real risk points, and more accurately characterize the vehicle's defensive emergency steering behavior after the potential risk points appear, thereby making a reasonable assessment of the autonomous driving capability and improving the safety of autonomous driving.

[0050] In combination with the foregoing first aspect, in a possible implementation manner, the fifth information further includes any one or more of the following:

[0051] duration of the straight-line driving of the first vehicle in the seventh interval, a peak steering wheel angle of the first vehicle during vehicle steering in the eighth interval, and a steering wheel rotation cycle of the first vehicle.

[0052] In an embodiment of the present application, a possible specific implementation of the fifth information is provided. Specifically, when the first information includes the second risk point, the fifth information required to construct the fourth reaction model may include, but is not limited to, the duration of the first vehicle's straight-line travel within the seventh interval, the peak steering wheel angle of the first vehicle during vehicle steering within the eighth interval, the steering wheel rotation period of the first vehicle, and so on. The fourth reaction model constructed based on the fifth information in the embodiment of the present application can more accurately characterize the vehicle's defenseless emergency steering behavior after the emergence of a potential risk point.

[0053] In a second aspect, an embodiment of the present application provides a method for calibrating parameters of a reaction model, which is applied to a first vehicle and can be performed by a communication device. The communication device can be a device, or a chip (system) or circuit for a device, which is not limited in this application. The method includes:

[0054] Acquiring sixth information, the sixth information including parameter information output by a reaction model, the parameter information being used to characterize driving behavior of a typical human driver corresponding to the first vehicle; wherein the reaction model is constructed based on the first information, the first information including information of a first risk point or a second risk point, the first risk point being a real risk point appearing on the road where the first vehicle is located at a first moment, the second risk point being a potential risk point appearing on the road where the first vehicle is located at the first moment that triggers a third risk point, the third risk point being a real risk point appearing on the road where the first vehicle is located within a first time period after the first moment, and the reaction model being used to characterize the first vehicle's action after the occurrence of the first risk point or the second risk point;

[0055] Target parameters are obtained based on the sixth information calibration, and the target parameters are used to evaluate the autonomous driving performance of the first vehicle.

[0056] In an embodiment of the present application, a parameter calibration method for a reaction model is provided, wherein a communication device obtains sixth information and calibrates target parameters based on the sixth information. The communication device here may also be a processor / chip that can be used to execute computer execution instructions, and the embodiment of the present application does not limit this.

[0057] The sixth information in the embodiment of the present application includes parameter information output by a reaction model. This parameter information is used to characterize the driving behavior of an average human driver corresponding to the first vehicle. The reaction model is constructed based on the first information and is used to characterize the first vehicle's behavior after the occurrence of a first risk point or a second risk point. The first information includes information about the first risk point or the second risk point. The first risk point is a real risk point that appears on the road where the first vehicle is located at a first moment. A real risk point can be understood as an event that occurs at the first moment and threatens the driving safety of the vehicle. The second risk point is a potential risk point that appears on the road where the first vehicle is located at a first moment and triggers a third risk point. The third risk point is a real risk point that appears on the road where the first vehicle is located within a first time period after the first moment. A potential risk point can be understood as an event that occurs at the first moment and may affect the driving safety of the vehicle but has not yet posed an actual threat to the driving safety of the vehicle, but the subsequent events it triggers pose a threat to the driving safety of the vehicle. It is understood that the second risk point can be used to provide a forewarning of a reasonably foreseeable real risk point within the first time period after the first moment. In the embodiment of the present application, the target parameters calibrated based on the sixth information are used to evaluate the autonomous driving performance of the first vehicle.

[0058] In an embodiment of the present application, target parameters are obtained by calibration based on the parameter information output by the reaction model, and are used to evaluate the autonomous driving performance of the first vehicle. Since the concept of potential risk points is introduced into the reaction model, more dangerous driving scenarios can be covered, and the definition logic of real risk points can be optimized, so as to more accurately characterize the vehicle's action behavior after the risk points appear. Therefore, the target parameters obtained by calibration based on the parameter information output by the reaction model can make a reasonable evaluation of the autonomous driving capability and improve the safety of autonomous driving.

[0059] Optionally, the reaction model in the second aspect can be constructed by the method described in the first aspect and any possible implementation method.

[0060] In conjunction with the above second aspect, in a possible implementation manner, when the reaction model includes the first reaction model or the second reaction model, the target parameter includes any one or more of the following:

[0061] The maximum braking deceleration of the first vehicle, the braking effectiveness improvement time of the first vehicle, the braking effectiveness improvement rate of the first vehicle, and the decision response time of an average human driver corresponding to the first vehicle;

[0062] The first reaction model is used to characterize the braking behavior of the first vehicle after the first risk point occurs, and the second reaction model is used to characterize the braking behavior of the first vehicle after the second risk point occurs.

[0063] In an embodiment of the present application, a possible specific implementation of a reaction model is provided, specifically, the reaction model used to calibrate the target parameters may include a first reaction model or a second reaction model, wherein the first reaction model is used to characterize the braking behavior of the first vehicle after the first risk point appears, and the second reaction model is used to characterize the braking behavior of the first vehicle after the second risk point appears. In this case, the target parameters obtained by calibration based on the parameter information output by the first reaction model or the second reaction model include but are not limited to the maximum braking deceleration of the first vehicle, the braking efficiency improvement time of the first vehicle, the braking efficiency improvement rate of the first vehicle, and the decision response time of the general human driver corresponding to the first vehicle. Through the embodiment of the present application, the target parameters obtained by calibration using the parameter information output by the first reaction model or the second reaction model can make a reasonable assessment of the capability of autonomous driving and improve the safety of autonomous driving.

[0064] In combination with the above-mentioned second aspect, in a possible implementation manner, the maximum braking deceleration of the first vehicle is obtained by calibrating N frames of data with larger braking deceleration in the parameter information after the first risk point or the second risk point occurs, where N is an integer greater than 0.

[0065] In an embodiment of the present application, a possible specific implementation of the maximum braking deceleration of the first vehicle is provided. Specifically, N frames (for example, 10 frames) of data with a larger braking deceleration after the first risk point or the second risk point occurs in the sliced ​​data can be selected and the average value can be taken as the maximum braking deceleration, which can more accurately characterize the braking behavior of the general human driver corresponding to the first vehicle after the first risk point or the second risk point occurs.

[0066] In combination with the above-mentioned second aspect, in a possible implementation manner, the braking efficiency improvement time of the first vehicle is the difference between the braking saturation moment and the braking start moment, the braking start moment is obtained by calibrating at least two frames of data with smaller braking deceleration in the parameter information, and the braking saturation moment is obtained by calibrating at least two frames of data with larger braking deceleration in the parameter information.

[0067] In an embodiment of the present application, a possible specific implementation of the braking performance improvement time of a first vehicle is provided. Specifically, the braking start time is determined by calibrating at least two frames of data with a relatively low braking deceleration in the parameter information output by the reaction model. The braking saturation time is determined by calibrating at least two frames of data with a relatively high braking deceleration in the parameter information output by the reaction model. The difference between the braking saturation time and the braking start time is the braking performance improvement time of the first vehicle. It is understood that a relatively low braking deceleration may refer to a braking deceleration less than a preset threshold. This preset threshold is not a fixed value and can be adjusted based on different application scenarios. For example, it can be two frames of data with braking decelerations of 10% and 15% of the maximum braking deceleration. Similarly, a relatively high braking deceleration may refer to a braking deceleration greater than a preset threshold. This preset threshold is not a fixed value and can be adjusted based on different application scenarios. For example, it can be two frames of data with braking decelerations of 90% and 85% of the maximum braking deceleration. In an embodiment of the present application, it is assumed that the braking efficiency improvement rate is the same at the beginning of braking and when it is about to be saturated. The braking saturation time is calculated by using the time when the braking deceleration is larger, and the braking start time is calculated by using the time when the braking deceleration is smaller. A robust and anti-interference braking efficiency improvement time can be obtained, thereby more accurately characterizing the braking behavior of the general human driver corresponding to the first vehicle after the first risk point or the second risk point occurs.

[0068] In combination with the above second aspect, in a possible implementation manner, the braking efficiency improvement rate of the first vehicle is obtained by calibrating at least one frame of data with a smaller braking deceleration and at least one frame of data with a larger braking deceleration in the parameter information.

[0069] In an embodiment of the present application, a possible specific implementation of the braking efficiency improvement rate of a first vehicle is provided. Specifically, the braking efficiency improvement rate of the first vehicle can be obtained by calibrating at least one frame of data with a low braking deceleration and at least one frame of data with a high braking deceleration in parameter information. It is understood that low braking deceleration can refer to a braking deceleration less than a preset threshold. This preset threshold is not a fixed value and can be adjusted according to different application scenarios. For example, it can be at least one frame of data with a braking deceleration of 10% of the maximum braking deceleration. Similarly, high braking deceleration can refer to a braking deceleration greater than a preset threshold. This preset threshold is not a fixed value and can be adjusted according to different application scenarios. For example, it can be at least one frame of data with a braking deceleration of 90% of the maximum braking deceleration. In this embodiment of the present application, by selecting the braking deceleration changes in the middle section of the first vehicle's entire braking process, a more accurate braking efficiency improvement rate of the first vehicle can be obtained, thereby more accurately characterizing the braking behavior of an average human driver of the first vehicle after the occurrence of the first risk point or the second risk point.

[0070] In combination with the above-mentioned second aspect, in a possible implementation manner, the decision response time of the general human driver corresponding to the first vehicle is the difference between the braking start time and the time when the actual risk point occurs, and the braking start time is calibrated by at least two frames of data with smaller braking deceleration in the parameter information.

[0071] In an embodiment of the present application, a possible specific implementation method for determining the decision response time of an average human driver corresponding to a first vehicle is provided. Specifically, the braking start time is calibrated from at least two frames of data with low braking deceleration in the parameter information. The difference between the braking start time and the time when the actual risk point occurs is used as the decision response time of the average human driver corresponding to the first vehicle. This can more accurately characterize the braking behavior of the average human driver corresponding to the first vehicle after the occurrence of the first or second risk point. It is understood that low braking deceleration can refer to a braking deceleration less than a preset threshold. This preset threshold is not a fixed value and can be adjusted according to different application scenarios. For example, it can be at least two frames of data with braking decelerations of 10% and 15% of the maximum braking deceleration.

[0072] In conjunction with the second aspect, in a possible implementation manner, when the reaction model includes the third reaction model or the fourth reaction model, the parameter information includes any one or more of the following:

[0073] the maximum steering wheel angle of the first vehicle, the steering wheel return time of the first vehicle, and the decision response time of an average human driver corresponding to the first vehicle;

[0074] Among them, the third reaction model is used to characterize the steering behavior of the first vehicle after the first risk point occurs, and the fourth reaction model is used to characterize the steering behavior of the first vehicle after the second risk point occurs.

[0075] In an embodiment of the present application, a possible specific implementation of a reaction model is provided, specifically, the reaction model used to calibrate the target parameters may include a third reaction model or a fourth reaction model, wherein the third reaction model is used to characterize the steering behavior of the first vehicle after the first risk point appears, and the fourth reaction model is used to characterize the steering behavior of the first vehicle after the second risk point appears. In this case, the target parameters obtained by calibration based on the parameter information output by the third reaction model or the fourth reaction model include but are not limited to the maximum steering wheel angle of the first vehicle, the steering wheel return time of the first vehicle, and the decision response time of the general human driver corresponding to the first vehicle. Through the embodiment of the present application, the target parameters obtained by calibration using the parameter information output by the third reaction model or the fourth reaction model can make a reasonable assessment of the capability of autonomous driving and improve the safety of autonomous driving.

[0076] In combination with the above-mentioned second aspect, in a possible implementation manner, the maximum steering wheel angle of the first vehicle is obtained by calibrating the M frames of data with a larger steering wheel angle in the parameter information after the first risk point or the second risk point occurs, where M is an integer greater than 0.

[0077] In an embodiment of the present application, a possible specific implementation of the maximum steering wheel angle of a first vehicle is provided. Specifically, M frames (for example, 10 frames) of data with a larger steering wheel angle after the first risk point or the second risk point occurs in the slice data can be selected and the average value can be taken as the maximum steering wheel angle, which can more accurately characterize the steering behavior of the general human driver corresponding to the first vehicle after the first risk point or the second risk point occurs.

[0078] In combination with the above-mentioned second aspect, in a possible embodiment, the steering wheel return time of the first vehicle is the difference between the first return time and the steering start time, the steering start time is obtained by calibrating at least two frames of data with smaller steering wheel angles in the parameter information, and the first return time is obtained by calibrating at least two frames of data with larger steering wheel angles in the parameter information.

[0079] In an embodiment of the present application, a possible specific implementation of the steering wheel return time of a first vehicle is provided. Specifically, at least two frames of data with a smaller steering wheel angle in the parameter information output by the reaction model are calibrated to obtain the turn start time, and at least two frames of data with a larger steering wheel angle in the parameter information output by the reaction model are calibrated to obtain the first return time. The difference between the first return time and the turn start time is the steering wheel return time of the first vehicle. It is understood that a smaller steering wheel angle can refer to a steering wheel angle less than a preset threshold. This preset threshold is not a fixed value and can be adjusted according to different application scenarios. For example, it can refer to two frames of data with steering wheel angles of 20% and 15% of the maximum steering wheel angle. Similarly, a larger steering wheel angle can refer to a steering wheel angle greater than a preset threshold. This preset threshold is not a fixed value and can be adjusted according to different application scenarios. For example, it can refer to two frames of data with steering wheel angles of 80% and 85% of the maximum steering wheel angle. In an embodiment of the present application, it is assumed that the steering angle increase rate is the same at the beginning of steering and when the first steering return is about to be made. The first steering return time is inferred from the moment at the larger steering angle, and the steering start time is inferred from the moment at the smaller steering angle. The robust and anti-interference steering wheel return time can be obtained, thereby more accurately characterizing the steering behavior of the general human driver corresponding to the first vehicle after the first risk point or the second risk point appears.

[0080] In combination with the above-mentioned second aspect, in a possible implementation manner, the decision response time of the general human driver corresponding to the first vehicle is the difference between the turning start time and the time when the real risk point appears, and the turning start time is calibrated by at least two frames of data with smaller steering wheel angles in the parameter information.

[0081] In an embodiment of the present application, a possible specific implementation method for the decision response time of an average human driver corresponding to a first vehicle is provided. Specifically, the steering start time is calibrated from at least two frames of data in the parameter information showing a small steering wheel angle. The difference between the steering start time and the time when the actual risk point occurs is used as the decision response time of the average human driver corresponding to the first vehicle. This can more accurately characterize the steering behavior of the average human driver corresponding to the first vehicle after the first or second risk point occurs. It is understood that a small steering wheel angle can refer to a steering wheel angle less than a preset threshold. This preset threshold is not a fixed value and can be adjusted according to different application scenarios. For example, it can be at least two frames of data showing steering wheel angles of 20% and 15% of the maximum steering wheel angle.

[0082] In a third aspect, an embodiment of the present application provides a communication device, which includes a module or unit for executing any method described in the first aspect.

[0083] In one possible design, the apparatus includes:

[0084] a processing unit configured to obtain first information, the first information including information of a first risk point or a second risk point, the first risk point being a real risk point occurring on the road where the first vehicle is currently located, the second risk point being a potential risk point occurring on the road where the first vehicle is currently located that may trigger a third risk point, the third risk point being a real risk point occurring on the road where the first vehicle is currently located within a first time period after the current moment;

[0085] The processing unit is further used to construct a reaction model based on the first information, and the reaction model is used to characterize the action behavior of the first vehicle after the first risk point or the second risk point occurs.

[0086] In one possible implementation, the device further includes:

[0087] a communication unit, configured to receive the first information;

[0088] The processing unit is specifically configured to obtain the first information through the communication unit.

[0089] The methods performed by the above-mentioned processing unit and communication unit can refer to the methods corresponding to the above-mentioned first aspect, and will not be repeated here.

[0090] Regarding the technical effects brought about by the third aspect and any possible implementation method, please refer to the introduction of the technical effects corresponding to the first aspect and the corresponding implementation method.

[0091] In a fourth aspect, an embodiment of the present application provides a communication device, which includes a module or unit for executing the method as described in any one of the second aspects.

[0092] In one possible design, the apparatus includes:

[0093] a processing unit configured to obtain sixth information, the sixth information including parameter information output by a reaction model, the parameter information being used to characterize driving behavior of a typical human driver corresponding to the first vehicle; wherein the reaction model is constructed based on the first information, the first information including information about a first risk point or a second risk point, the first risk point being a real risk point appearing on the road where the first vehicle is located at a first moment, the second risk point being a potential risk point appearing on the road where the first vehicle is located at a first moment that triggers a third risk point, the third risk point being a real risk point appearing on the road where the first vehicle is located within a first time period after the first moment, and the reaction model being used to characterize the first vehicle's action after the occurrence of the first risk point or the second risk point;

[0094] The processing unit is further used to calibrate and obtain target parameters based on the sixth information, and the target parameters are used to evaluate the autonomous driving performance of the first vehicle.

[0095] In one possible implementation, the device further includes:

[0096] a communication unit, configured to receive the sixth information;

[0097] The processing unit is specifically configured to obtain the sixth information through the communication unit.

[0098] The methods performed by the above-mentioned processing unit and communication unit can refer to the methods corresponding to the above-mentioned second aspect, and will not be repeated here.

[0099] Regarding the technical effects brought about by the fourth aspect and any possible implementation method, please refer to the introduction of the technical effects corresponding to the second aspect and the corresponding implementation method.

[0100] Optionally, in the communication device described in any one of the third to fourth aspects and any possible implementation manner:

[0101] In one implementation, the communication apparatus is a communication device. When the communication apparatus is a communication device, the communication unit may be a transceiver or an input / output interface; and the processing unit may be at least one processor. Alternatively, the transceiver may be a transceiver circuit. Alternatively, the input / output interface may be an input / output circuit.

[0102] In another implementation, the communication device is a chip (system) or circuit used in a communication device. When the communication device is a chip (system) or circuit used in a communication device, the communication unit may be a communication interface (input / output interface), interface circuit, output circuit, input circuit, pin, or related circuit on the chip (system) or circuit; and the processing unit may be at least one processor, processing circuit, or logic circuit.

[0103] In a fifth aspect, embodiments of the present application provide a communication device comprising a processor. The processor is coupled to a memory and can be configured to execute instructions in the memory to implement the method of any of the first and second aspects described above and any possible implementation methods. Optionally, the communication device further comprises a memory. Optionally, the communication device further comprises a communication interface, the processor being coupled to the communication interface.

[0104] In a sixth aspect, embodiments of the present application provide a communication device, comprising: a logic circuit and a communication interface. The communication interface is configured to receive or send information; the logic circuit is configured to receive or send information via the communication interface, so that the communication device executes the method of any of the first and second aspects above, and any possible implementation thereof.

[0105] In the seventh aspect, an embodiment of the present application provides a computer-readable storage medium, which is used to store a computer program (also referred to as code, or instructions); when the computer program is run on a computer, the method of any one of the above-mentioned first to second aspects and any possible implementation method is implemented.

[0106] In an eighth aspect, an embodiment of the present application provides a computer program product, comprising: a computer program (also referred to as code, or instructions); when the computer program is run, it enables the computer to execute any one of the above-mentioned first to second aspects and any possible implementation method.

[0107] In a ninth aspect, an embodiment of the present application provides a chip, comprising a processor configured to execute instructions. When the processor executes the instructions, the chip performs the method of any one of the first and second aspects and any possible implementation methods described above. Optionally, the chip further comprises a communication interface configured to receive or send signals.

[0108] In the tenth aspect, an embodiment of the present application provides a vehicle end, which includes at least one communication device as described in the third aspect, or the communication device as described in the fourth aspect, or the communication device as described in the fifth aspect, or the communication device as described in the sixth aspect, or the chip as described in the ninth aspect.

[0109] In addition, in the process of executing the method described in any aspect of the first aspect to the second aspect and any possible implementation method, the process of sending information and / or receiving information in the above method can be understood as the process of the processor outputting information and / or the process of the processor receiving input information. When outputting information, the processor can output the information to the transceiver (or communication interface, or sending module) so that it can be transmitted by the transceiver. After the information is output by the processor, it may also need to undergo other processing before it reaches the transceiver. Similarly, when the processor receives input information, the transceiver (or communication interface, or sending module) receives the information and inputs it into the processor. Furthermore, after the transceiver receives the information, the information may need to undergo other processing before it is input into the processor.

[0110] Based on the above principles, for example, the sending of information mentioned in the above method can be understood as the processor outputting information. For another example, the receiving of information can be understood as the processor receiving input information.

[0111] Optionally, for the operations such as transmission, sending and receiving involved in the processor, if there is no special explanation, or if they do not conflict with their actual functions or internal logic in the relevant description, they can be more generally understood as processor output, reception, input and other operations.

[0112] Optionally, in the process of executing the method described in any aspect of the first to second aspects and any possible implementation method, the processor may be a processor specifically used to execute these methods, or a processor that executes these methods by executing computer instructions in a memory, such as a general-purpose processor. The memory may be a non-transitory memory, such as a read-only memory (ROM), which may be integrated with the processor on the same chip or may be separately provided on different chips. The embodiments of the present application do not limit the type of memory and the configuration of the memory and the processor.

[0113] In a possible implementation, the at least one memory is located outside the device.

[0114] In yet another possible implementation, the at least one memory is located within the device.

[0115] In another possible implementation, part of the at least one memory is located inside the device, and another part of the memory is located outside the device.

[0116] In this application, the processor and the memory may also be integrated into one device, that is, the processor and the memory may also be integrated together.

[0117] In the embodiments of the present application, by introducing the concept of potential risk points, more dangerous driving scenarios can be covered, and the definition logic of real risk points can be optimized, so as to more accurately characterize the vehicle's action behavior after the risk points appear, thereby making a reasonable assessment of the autonomous driving capabilities and improving the safety of autonomous driving. BRIEF DESCRIPTION OF THE DRAWINGS

[0118] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following briefly introduces the drawings required for use in the embodiments of the present application. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0119] FIG1 is a schematic diagram of a driver reaction model provided in an embodiment of the present application;

[0120] FIG2 is a schematic flow chart of a reaction model construction method provided in an embodiment of the present application;

[0121] FIG3A is a schematic diagram of a vehicle cutting-in scenario provided by an embodiment of the present application;

[0122] FIG3B is a schematic diagram of a vehicle cutting-in scenario provided by an embodiment of the present application;

[0123] FIG3C is a schematic diagram of a vehicle cutting-in scenario provided in an embodiment of the present application;

[0124] FIG4A is a schematic diagram of a scene in which a small vehicle cuts in according to an embodiment of the present application;

[0125] FIG4B is a schematic diagram of a scene of a large vehicle cutting in according to an embodiment of the present application;

[0126] FIG4C is a schematic diagram of a scene in which a leading vehicle cuts out according to an embodiment of the present application;

[0127] FIG5 is a schematic diagram of a driver braking and steering model provided in an embodiment of the present application;

[0128] FIG6 is a flow chart of a parameter calibration method for a reaction model provided in an embodiment of the present application;

[0129] FIG7 is a schematic structural diagram of a communication device provided in an embodiment of the present application;

[0130] FIG8 is a schematic structural diagram of a communication device provided in an embodiment of the present application;

[0131] FIG9 is a schematic diagram of the structure of a chip provided in an embodiment of the present application. DETAILED DESCRIPTION

[0132] In order to make the purpose, technical solutions and advantages of this application clearer, the embodiments of this application will be described below in conjunction with the drawings in the embodiments of this application.

[0133] The terms "first" and "second" in the specification, claims, and drawings of this application are used to distinguish different objects, not to describe a specific order. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product, or device that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units that are not listed, or may optionally include other steps or units that are inherent to the process, method, product, or device.

[0134] The “embodiment” mentioned herein means that the specific features, structures or characteristics described in conjunction with the embodiment may be included in at least one embodiment of the present application. The appearance of this phrase in various places in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It can be understood by those skilled in the art that, in the various embodiments of the present application, unless otherwise specified and there is a logical conflict, the terms and / or descriptions between the various embodiments are consistent and can be referenced to each other, and the technical features in different embodiments can be combined to form a new embodiment according to their inherent logical relationships.

[0135] It should be understood that in the present application, "at least one (item)" refers to one or more, "more than one" refers to two or more, "at least two (items)" refers to two or three and more than three, and "and / or" is used to describe the association relationship of associated objects, indicating that three relationships may exist. For example, "A and / or B" can mean: only A exists, only B exists, and A and B exist at the same time, where A and B can be singular or plural. The character " / " generally indicates that the previous and next associated objects are in an "or" relationship. "At least one of the following items" or similar expressions refers to any combination of these items, including any combination of single or plural items. For example, at least one of a, b or c can mean: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, c can be single or multiple.

[0136] It should be noted that in this application, "indication" can include direct indication, indirect indication, explicit indication, and implicit indication. When describing that a certain indication information is used to indicate A, it can be understood that the indication information carries A, directly indicates A, or indirectly indicates A.

[0137] In this application, the information indicated by the indication information is referred to as the information to be indicated. In specific implementations, there are many ways to indicate the information to be indicated. For example, but not limited to, the information to be indicated can be directly indicated, such as the information to be indicated itself or an index of the information to be indicated. The information to be indicated can also be indirectly indicated by indicating other information, where the other information is associated with the information to be indicated. Alternatively, only a portion of the information to be indicated can be indicated, while the rest of the information to be indicated is known or agreed upon in advance. For example, the indication of specific information can be achieved by using a pre-agreed (e.g., protocol-specified) order of the various information, thereby reducing indication overhead to a certain extent. The information to be indicated can be sent as a whole or divided into multiple sub-information and sent separately. The transmission period and / or transmission timing of these sub-information can be the same or different. The specific transmission method is not limited in this application. The transmission period and / or transmission timing of these sub-information can be pre-defined, for example, according to a protocol, or can be configured by the transmitting device sending configuration information to the receiving device.

[0138] It should be noted that in this application, "send" can be understood as "output" and "receive" can be understood as "input". "Send information to A", where "to A" only indicates the direction of information transmission, A is the destination, and does not limit "sending information to A" to direct transmission on the air interface. "Sending information to A" includes sending information directly to A, and also includes sending information indirectly to A through a transmitter, so "sending information to A" can also be understood as "outputting information to A". Similarly, "receiving information from A" indicates that the source of the information is A, including receiving information directly from A, and also including receiving information indirectly from A through a receiver, so "receiving information from A" can also be understood as "inputting information from A".

[0139] This application provides a reaction model construction method and a reaction model parameter calibration method, which are applied to the field of intelligent driving technology, such as the evaluation of autonomous driving capabilities. In order to more clearly describe the solution of this application, some knowledge related to autonomous driving is first introduced below.

[0140] Autonomous driving technology is based on artificial intelligence (AI). It uses sensors, cameras, radar, lasers, and other sensing devices to perceive and interpret the environment, enabling autonomous driving and significantly improving driving efficiency. The industry consensus on acceptable autonomous driving capabilities is that autonomous driving must avoid reasonably foreseeable and avoidable accidents within its designed operating range.

[0141] The commonly used reference model for "reasonably foreseeable" is a semantic-level quantification of common driving scenarios. This model collects data from actual road conditions, statistically calculates the parameter distribution of logical scenarios, and uses the concept of "low-probability events" to determine the reasonably foreseeable parameter range. The commonly used reference model for "avoidability" is the reaction model of a careful and competent human driver (CC Driver). This reaction model can be used to assess whether a possible accident in a driving scenario is avoidable for a typical human driver, thereby evaluating the capabilities of autonomous driving.

[0142] If autonomous driving can achieve the above-mentioned level of risk avoidance of an average driver under reasonably foreseeable scenarios, it can be determined that the autonomous driving meets the driving behavior safety criteria.

[0143] Currently, the evaluation criteria used for automated lane keeping systems (ALKS) are that the autonomous driving system cannot bring about foreseeable and avoidable risks (accidents) during its operation and meets local traffic regulations (except in cases where violations are required to avoid accidents). The "avoidable" risk boundary is defined through the reaction model of ordinary human drivers, and the "foreseeable" accident scenario is defined through low-probability events in the logical scenario parameter space.

[0144] ALKS's interaction scenarios are categorized as avoidable and unavoidable. Thresholds for both scenarios are set based on simulated, attentive driver responses. Optionally, the operating speed can be set below 60 km / h. Under these conditions, the human driver's collision avoidance ability is primarily characterized by braking behavior. Therefore, an appendix proposes standard braking models applicable to three driving logic scenarios: adjacent vehicle cutting in, preceding vehicle cutting out, and preceding vehicle braking.

[0145] Specifically, the general human driver's reaction model is divided into three stages: perception, decision, and reaction. Accordingly, the parameters that need to be calibrated in this driver reaction model include the risk perception point, perception time, decision delay, maximum braking deceleration, and braking efficiency improvement time.

[0146] However, the driver reaction model still has many parameters that are difficult to determine and standardize, including but not limited to the following:

[0147] (1) Regarding the point at which risk perception occurs, the judgment methods for different scenarios are not uniform and do not conform to the driving cognition of ordinary human drivers. Taking the neighboring vehicle cutting in as an example, the point at which risk perception occurs is when the lateral offset of the neighboring vehicle exceeds the median of the lateral swing of natural driving in the center. If the neighboring vehicle has already turned on the turn signal in advance, the point at which risk perception occurs should be the time when the neighboring vehicle begins to change its lateral position. The point at which risk perception occurs in the scenario where the leading vehicle cuts out is still calculated based on the swing amount, but the real risk of this scenario lies in the slow speed of the leading vehicle in front of the cutting-out vehicle. Therefore, the point at which risk perception occurs in this scenario should be related to the appearance of the leading vehicle in front of the cutting-out vehicle, and has no direct relationship with the cutting-out action of the cutting-out vehicle.

[0148] (2) The risk perception time and decision-making time for different scenarios are consistent, which is inconsistent with the driving cognition of ordinary human drivers. Taking the turn signal cut-in scenario as an example, assuming that under the same cut-in scenario parameter settings, one neighboring car does not turn on the turn signal, and the other turns on the turn signal in advance, the driver's reaction time in the above two scenarios will be obviously different from the neighboring car's cut-in action to the driver's stepping on the brake pedal. If the cut-in vehicle is a large vehicle, the driver may have released the accelerator pedal before the neighboring car has made any obvious cut-in action, and the reaction time will be more timely.

[0149] (3) The current reaction model only provides the driver's braking reaction logic. From pedal analysis to control application, the model cannot directly solve the vehicle's operating state at the corresponding moment, and cannot uniquely quantify the parameters of the scenario that needs to be avoided. There is also the problem of not being able to directly solve the actual braking efficiency analysis, which is related to the road adhesion coefficient, tire temperature, and brake pressure build-up time, and is not solely determined by the brake pedal opening.

[0150] Therefore, there are many parameters mentioned above that are difficult to determine and standardize, resulting in the driver reaction model still being unable to cover most dangerous driving scenarios, thereby making an unreasonable assessment of the autonomous driving capability and affecting the safety of autonomous driving.

[0151] In view of this, this application designs a method for quickly building a reaction model of a focused average human driver (CC Driver) in different autonomous driving scenarios, as well as a parameter calibration method for the reaction model. It introduces the concept of potential risk points, optimizes the definition logic of real risk points, and integrates the two action logics of steering and braking to fully and truly reflect the differences in the defensive and stress responses of average human drivers. The reaction model obtained by the model construction method in this application can cover more dangerous driving scenarios and more accurately characterize the vehicle's behavior after the risk point appears, thereby making a reasonable assessment of the autonomous driving capability and improving the safety of autonomous driving.

[0152] The reaction model construction method and the reaction model parameter calibration method provided in the embodiments of the present application are described below with reference to Figures 1 to 5.

[0153] Please refer to FIG1 , which is a schematic diagram of a driver reaction model provided in an embodiment of the present application.

[0154] As shown in FIG1 , the driver reaction model is a research route for driver models used in evaluating autonomous driving capabilities.

[0155] First, the driver response model divides the standard human driver's actions into two actions and two phases. The two actions are steering and braking. The braking action can be further divided into releasing the accelerator pedal, moving the pedal, and pressing the brake pedal. The two phases are the defensive phase and the stress phase.

[0156] Secondly, based on the distinction between the two actions and the two stages, we can obtain a suitable action combination model for differentiated scenarios, as follows:

[0157] For reasonably foreseeable driving scenarios without warning (i.e., no potential risk points), the system first determines when a true risk point will appear (e.g., a neighboring vehicle's obvious intention to cut in), followed by risk identification, risk assessment, and decision-making. In low-speed driving scenarios, emergency braking is employed, i.e., a non-defensive emergency braking model after a true risk point appears. In high-speed driving scenarios with evasive space, emergency steering is employed, i.e., a non-defensive emergency steering model after a true risk point appears.

[0158] For reasonably foreseeable driving scenarios with foreshadowing (i.e., potential risk points), the first step is to determine when the potential risk will occur (for example, an oncoming vehicle has activated its turn signal in advance, there is a crosswalk ahead, a school bus has activated its stop sign, etc.). Secondly, determine whether there is room for lateral maneuvering. If so, defensive steering is employed to increase lateral distance, i.e., the vehicle's defensive emergency steering model after the potential risk point appears. If no lateral maneuvering space exists, the vehicle speed is reduced to a reasonable speed, i.e., the vehicle's defensive emergency braking model after the potential risk point appears. The defensive action model can complete relevant actions before the actual risk point appears, placing the vehicle in a relatively safe and stable state. When the actual risk occurs, due to psychological expectations, a shorter risk identification, risk assessment, and decision-making time can be used to implement reasonable steering or braking actions.

[0159] Based on the classification of the two actions and two stages of standard operations of general human drivers by the reaction model in Figure 1 above, the embodiment of the present application correspondingly provides a method for constructing the reaction model.

[0160] Please refer to Figure 2, which is a flow chart of a reaction model construction method provided in an embodiment of the present application. The reaction model construction method is applied to the field of intelligent driving technology, such as the evaluation of autonomous driving capabilities. The reaction model construction method is specifically applied to the first vehicle and includes but is not limited to the following steps:

[0161] S201: Acquire first information.

[0162] S202: Construct a reaction model based on the first information.

[0163] It is understood that the communication device used to obtain the first information and construct the reaction model based on the first information in the embodiment of the present application can be a device equipped with a processor / chip that can be used to execute computer-executable instructions, or a processor / chip that can be used to execute computer-executable instructions. The embodiment of the present application does not limit this. Optionally, the communication device can be a processor / chip in a smart car, which is used to execute the reaction model construction method in the embodiment of the present application to achieve coverage of more dangerous driving scenarios and more accurately characterize the vehicle's behavior after the risk point appears, thereby making a reasonable assessment of the ability of autonomous driving and improving the safety of autonomous driving.

[0164] The first information mentioned above includes information about the first risk point or the second risk point.

[0165] The first risk point is a real risk point that appears on the road where the first vehicle is located at the current moment. The real risk point can be understood as an event that threatens the driving safety of the vehicle and occurs at the current moment.

[0166] The second risk point is a potential risk point that appears on the road where the first vehicle is located at the current moment and triggers the third risk point. The third risk point is a real risk point that appears on the road where the first vehicle is located within the first time period after the current moment. The potential risk point can be understood as an event that occurs at the current moment and may affect the vehicle driving safety, but has not yet posed an actual threat to the vehicle driving safety. The subsequent events triggered by it pose a threat to the vehicle driving safety.

[0167] It is understandable that the second risk point can be used to provide a foreshadowing warning of a reasonably foreseeable real risk point within a first time period after the current moment. The first time period is not a fixed value and can be adjusted according to different application scenarios.

[0168] In the embodiment of the present application, the reaction model constructed based on the above-mentioned first information is used to characterize the action behavior of the first vehicle after the above-mentioned first risk point or the above-mentioned second risk point occurs.

[0169] In a possible embodiment, the second risk point may include, but is not limited to, any one or more of the following:

[0170] There are sections of the road where the first vehicle is located that have deceleration warnings, such as tunnel entrances, sharp bends, construction sections, sections at the top of ramps that affect safe sight distance, deceleration lanes, ramps, downhill sections, etc.;

[0171] The first vehicle is on a road with warning signs, such as warning signs for falling rocks, crosswinds, slippery conditions, wild animals, and high road bumps.

[0172] The first vehicle is on a road section with a static blind spot. For example, when overtaking a vehicle in an adjacent lane, there is a static blind spot in front of the vehicle being overtaken, and a pedestrian may cross the road.

[0173] A vehicle in an adjacent lane to the lane where the first vehicle is located brakes;

[0174] The adjacent vehicle of the first vehicle is a large vehicle;

[0175] The adjacent vehicles of the first vehicle correctly turned on the corresponding warning signals in advance before taking action, etc.

[0176] It is understandable that the above-mentioned potential risk situations can be used to provide advance warnings for reasonably foreseeable real risk points within the first time period after the current moment. By introducing the concept of potential risk points, more dangerous driving scenarios can be covered.

[0177] In a possible embodiment, the action behavior of the first vehicle after the occurrence of the first risk point or the second risk point represented by the reaction model may include but is not limited to at least one of vehicle braking behavior or vehicle steering behavior.

[0178] It is understandable that this reaction model integrates the two action logics of steering and braking, comprehensively and truly reflects the differences in defensive and stress responses of ordinary human drivers, and can more accurately characterize the vehicle's action behavior after the risk point appears.

[0179] In the embodiment of the present application, a method is designed to quickly construct a reaction model of a general human driver with concentrated attention in different autonomous driving scenarios. By introducing the concept of potential risk points, more dangerous driving scenarios can be covered, and the definition logic of real risk points can be optimized to more accurately characterize the vehicle's action behavior after the risk points appear, thereby making a reasonable assessment of the autonomous driving capabilities and improving the safety of autonomous driving.

[0180] The following describes the determination of risk points in the adjacent vehicle cut-in scenario with reference to FIG. 3A , FIG. 3B , and FIG. 3C .

[0181] Please refer to Figures 3A, 3B and 3C, which are schematic diagrams of vehicle entry scenarios provided in an embodiment of the present application.

[0182] FIG3A shows a scenario in which a vehicle has a slight lateral deviation when performing lane keeping.

[0183] FIG3B is a diagram illustrating the bilateral lateral offset range when a vehicle performs lane keeping.

[0184] FIG3C is a diagram showing a scenario where a neighboring vehicle cuts into a risk point in the scene.

[0185] Specifically, the neighboring vehicle cut-in scenario is defined as an adjacent vehicle in the adjacent lane traveling in the same direction as the main vehicle, switching lanes to the front of the main vehicle in the same lane, with no other vehicles between them at the time of the lane change. The moment the neighboring vehicle begins changing lanes indicates that a risk has emerged in the direction of travel for the main vehicle, so the moment the neighboring vehicle explicitly begins changing lanes is the risk point for this scenario.

[0186] The time point when a vehicle starts to change lanes can be determined by the lateral offset of the vehicle relative to the center line of its lane. Only when the lateral offset reaches a certain value can it be determined that the vehicle is changing lanes. This is because when a vehicle performs lane keeping behavior in its lane, it will inevitably deviate slightly from the center line of the lane and "wander" slightly in the lane, as shown in Figure 3A. Therefore, it is only necessary to count the lateral offset range of the vehicle center when the vehicle performs normal lane keeping behavior, as shown in Figure 3B. 1 / 2 of this range can be considered as the boundary point where the vehicle is about to perform lane change rather than lane keeping, which is the risk point of the above-mentioned neighboring vehicle cutting in scenario. Optionally, when the lateral offset range is 0.820m, the risk point of the neighboring vehicle cutting in scenario is that the neighboring vehicle deviates from the lane center line by 0.410m, as shown in Figure 3C, that is, when the neighboring vehicle deviates from the lane center line by more than 0.410m, it can be considered as the risk point of the scenario.

[0187] In a possible embodiment, the determination of the first risk point (real risk point) and / or the second risk point (potential risk point) in the reaction model construction method in Figure 2 can be explained respectively based on the method for determining the risk points in the neighboring vehicle cutting-in scenario in Figures 3A to 3C above, and in combination with the various driving scenarios shown in Figures 4A, 4B and 4C.

[0188] Please refer to FIG4A , which is a schematic diagram of a scene of a small vehicle cutting in according to an embodiment of the present application.

[0189] As shown in Figure 4A, in a scenario where a second vehicle cuts in front of a first vehicle (the vehicle itself) without turning on the turn signal, and the model of the second vehicle is not larger than that of the first vehicle, if the offset between the second vehicle and the center line of the lane in which the second vehicle is located is greater than the first threshold, it is determined as the first risk point.

[0190] The first threshold is not a fixed value and can be adjusted according to different application scenarios, which is not limited in the present embodiment. Optionally, the first threshold can be 0.410 m as determined in FIG. 3C .

[0191] The model of the second vehicle is not larger than that of the first vehicle, which can be understood as the second vehicle being a small vehicle. The current method for determining the first risk point is applicable to the small vehicle entry scenario.

[0192] It can be understood that in the scenario in which the second vehicle in the embodiment of the present application cuts in front of the first vehicle without turning on the turn signal, the second vehicle does not turn on the turn signal, is not a large vehicle, and there is no braking action. Therefore, it can be considered that there is no potential risk point in this scenario, that is, it does not pose a potential risk to the driving safety of the first vehicle. The first risk point at this time may be the moment when the second vehicle shows obvious intention to cut in.

[0193] Accordingly, the reaction model constructed based on the first information is used to characterize the defenseless emergency action behavior of the first vehicle in the scenario after the first risk point appears.

[0194] Please refer to FIG4B , which is a schematic diagram of a scene of a large vehicle cutting in according to an embodiment of the present application.

[0195] As shown in Figure 4B, in the scenario where a third vehicle cuts in front of the first vehicle without turning on the turn signal, if the model of the third vehicle is larger than that of the first vehicle, the third vehicle can be determined as the second risk point before the third vehicle cuts in. If the third vehicle deviates laterally, it will be determined as the third risk point.

[0196] Among them, the model of the third vehicle is larger than that of the first vehicle, which can be understood as the third vehicle being a large vehicle. The current method for determining the second risk point is applicable to the scenario where a large vehicle enters.

[0197] It is understandable that in the scenario described in the embodiment of the present application where a third vehicle cuts in front of the first vehicle without turning on its turn signal, although the third vehicle does not turn on its turn signal, it is a large vehicle adjacent to the first vehicle. Therefore, it can be considered that there is a potential risk point in this scenario, that is, it poses a potential risk to the driving safety of the first vehicle. In this case, the second risk point can be determined to be the large vehicle adjacent to the first vehicle (i.e., the third vehicle). Because the third vehicle is a large vehicle, its cutting-in action is more obvious. Therefore, the lateral deviation of the third vehicle can be determined as the third risk point, that is, the actual risk point that may be caused by the second risk point.

[0198] It is understandable that the second risk point (the large vehicle adjacent to the first vehicle) can be used to provide a foreshadowing warning of a real risk point (lateral deviation of the third vehicle) that is reasonably foreseeable within the first time period after the current moment.

[0199] Accordingly, the reaction model constructed based on the first information is used to characterize the defensive emergency action behavior of the first vehicle in the scenario after the second risk point appears.

[0200] Please refer to FIG. 4C , which is a schematic diagram of a scene in which a leading vehicle cuts out according to an embodiment of the present application.

[0201] As shown in Figure 4C, when the fifth vehicle in front of the first vehicle cuts out of the current lane, the first vehicle encounters the fourth vehicle in front of the fifth vehicle. If the lateral body of the fourth vehicle appearing in the field of view of the first vehicle is greater than the second threshold, it is determined as the first risk point.

[0202] The second threshold is not a fixed value and can be adjusted according to different application scenarios, which is not limited in the embodiment of the present application.

[0203] It can be understood that in the scenario in which the fifth vehicle cuts out of the current lane and the first vehicle encounters the fourth vehicle in front of the fifth vehicle in the embodiment of the present application, there is no potential risk point, that is, it does not pose a potential risk to the driving safety of the first vehicle. The first risk point at this time can be the moment when the lateral body of the fourth vehicle that appears in the field of view of the first vehicle is greater than the second threshold.

[0204] Accordingly, the reaction model constructed based on the first information is used to characterize the defenseless emergency action behavior of the first vehicle in the scenario after the first risk point appears.

[0205] It should be understood that the above Figures 4A, 4B, and 4C are merely exemplary illustrations of determining the first risk point (real risk point) and / or the second risk point (potential risk point) under three possible driving scenarios, and should not be used to limit the embodiments of the present application. Embodiments obtained based on the above driving scenarios and supplements or reasonable deformations of the determination of the first risk point (real risk point) and / or the second risk point (potential risk point) all fall within the scope of protection of the embodiments of the present application.

[0206] In a possible embodiment, the step S202 of "constructing a reaction model based on the first information" in FIG. 2 may be further described in combination with the driver braking and steering model shown in FIG. 5 , with specific examples as follows:

[0207] Case 1:

[0208] In a case where the first information includes a first risk point, second information is determined, and a first reaction model is constructed based on the first information and the second information.

[0209] When the first information includes a first risk point, it indicates that the road where the first vehicle is currently located is a real risk point. In this case, the determined second information includes information about a first section where the first vehicle is traveling at a constant speed and information about a second section where the vehicle is braking.

[0210] Correspondingly, the first reaction model constructed based on the first information and the second information is used to characterize the defenseless emergency braking behavior of the first vehicle in this scenario after the first risk point appears, that is, after the first risk point appears, the first vehicle continues to travel at a constant speed in the first interval, and then starts to brake the vehicle in the second interval, and reaches the maximum braking deceleration with a certain slope, and keeps the vehicle decelerating until it stops.

[0211] It can be understood that when the first information includes the first risk point, the second information required to construct the first reaction model may include but is not limited to: the duration of the first vehicle's uniform speed driving in the first interval, the starting time point of the first vehicle's vehicle braking in the second interval, the braking rate of the first vehicle, the maximum braking deceleration of the first vehicle, and the like.

[0212] The first reaction model constructed based on the second information in the embodiment of the present application can more accurately characterize the vehicle's defenseless emergency braking behavior after the real risk point occurs.

[0213] Please refer to FIG5 for details, which is a schematic diagram of a driver braking and steering model provided in an embodiment of the present application.

[0214] As shown in FIG5 , the non-defensive emergency braking model in FIG5 can be divided into three areas: free driving area, uniform speed area, and mechanical braking area.

[0215] In the free-running zone, the driver drives according to their driving needs. After the first risk point appears, the vehicle enters the uniform speed zone (the first interval mentioned above), where the driver conducts risk assessment, makes decisions, moves the foot pedal, and begins to apply the brake pedal. In the mechanical braking zone (the second interval mentioned above), the vehicle experiences a significant braking effect. After a rapid increase in braking efficiency, the vehicle reaches maximum deceleration and maintains deceleration until it stops.

[0216] It can be seen that in the non-defensive emergency braking model in Figure 5, the parameters to be confirmed for constructing the non-defensive emergency braking model may include but are not limited to: the time point of risk occurrence, the duration of the uniform speed period, the industry average braking delay, the industry average pressure building time, the industry average maximum braking deceleration, the braking start time point, the braking opening slope, the maximum braking deceleration, and so on.

[0217] The first reaction model constructed through the embodiments of the present application can cover driving scenarios where real risk points appear, optimize the definition logic of real risk points, and more accurately characterize the vehicle's defenseless emergency braking behavior after the real risk points appear, thereby making a reasonable assessment of the autonomous driving capabilities and improving the safety of autonomous driving.

[0218] Case 2:

[0219] In the case where the first information includes the second risk point, third information is determined, and a second reaction model is constructed based on the first information and the third information.

[0220] When the first information includes a second risk point, it indicates that the road currently occupied by the first vehicle is a potential risk point. This can be used to provide a foreshadowing warning of a third risk point (i.e., a real risk point) that is reasonably foreseeable within a first time period after the current moment. In this case, the determined third information includes information about the third section in which the first vehicle is traveling at a constant speed and information about the fourth section in which the vehicle is braking.

[0221] Correspondingly, the second reaction model constructed based on the first information and the third information is used to characterize the defensive emergency braking behavior of the first vehicle in this scenario after the second risk point appears, that is, after the second risk point appears and before the third risk point appears, the first vehicle maintains a gliding state, that is, it drives in the current driving state. After the third risk point appears, the first vehicle continues to drive at a constant speed in the third interval, and then starts to brake the vehicle in the fourth interval, and reaches the maximum braking deceleration at a certain slope, and keeps the vehicle decelerating until it stops.

[0222] It can be understood that when the first information includes the second risk point, the third information required to construct the second reaction model may include but is not limited to: the duration of the first vehicle's uniform speed driving in the third interval, the starting time point of the first vehicle's vehicle braking in the fourth interval, the braking rate of the first vehicle, the maximum braking deceleration of the first vehicle, and the like.

[0223] The second reaction model constructed based on the third information in the embodiment of the present application can more accurately characterize the defensive emergency braking behavior of the vehicle after a potential risk point appears.

[0224] Optionally, the second reaction model constructed in this embodiment is similar to the non-defensive emergency braking model in Figure 5 above. The difference between the two is that the third interval in the second reaction model is different from the uniform speed driving area (i.e., the above-mentioned first interval) in the non-defensive emergency braking model in Figure 5 above. Specifically, the starting time point, ending time point, interval length, etc. of the interval are different. The two intervals are respectively suitable for defensive braking scenarios and stress braking scenarios; the fourth interval in the second reaction model is different from the mechanical braking area (i.e., the above-mentioned second interval) in the non-defensive emergency braking model in Figure 5 above. Specifically, the starting time point, ending time point, interval length, etc. of the interval are different. The two intervals are respectively suitable for defensive braking scenarios and stress braking scenarios.

[0225] The second reaction model constructed through the embodiment of the present application introduces the concept of potential risk points, which can cover driving scenarios where potential risk points appear, and can optimize the definition logic of real risk points, and more accurately characterize the vehicle's defensive emergency braking behavior after the potential risk points appear, thereby making a reasonable assessment of the autonomous driving capability and improving the safety of autonomous driving.

[0226] Case 3:

[0227] In a case where the first information includes a first risk point, fourth information is determined, and a third reaction model is constructed based on the first information and the fourth information.

[0228] When the first information includes the first risk point, it indicates that the road where the first vehicle is currently located is a real risk point. In this case, the fourth information determined includes information about the fifth section where the first vehicle is traveling in a straight line and information about the sixth section where the vehicle is turning.

[0229] Correspondingly, the third reaction model constructed based on the first information and the fourth information is used to characterize the defenseless emergency turning behavior of the first vehicle in this scenario after the first risk point appears, that is, after the first risk point appears, the first vehicle continues to drive in a straight line in the fifth interval, and then starts to turn the vehicle in the sixth interval, first reaching the maximum turning angle in a sinusoidal form and then returning to the positive turning angle, and continues to drive in a straight line while maintaining the direction after returning to the positive turning angle.

[0230] It can be understood that when the first information includes the first risk point, the fourth information required to construct the third reaction model may include but is not limited to: the duration of the first vehicle's straight-line driving in the fifth interval, the peak steering wheel angle of the first vehicle when turning the vehicle in the sixth interval, the steering wheel rotation cycle of the first vehicle, and so on.

[0231] The third reaction model constructed based on the fourth information in the embodiment of the present application can more accurately characterize the vehicle's defenseless emergency steering behavior after the real risk point appears.

[0232] Please refer to FIG5 for details, which is a schematic diagram of a driver braking and steering model provided in an embodiment of the present application.

[0233] As shown in FIG5 , the non-defensive emergency steering model in FIG5 can be divided into two areas: a straight-line driving area and a sinusoidal steering area.

[0234] In the straight driving zone, the driver drives according to driving requirements. After the first risk point appears, while the vehicle is still in the straight driving state (i.e., the fifth interval mentioned above), the driver conducts risk assessment, makes decisions, and begins to grip and turn the steering wheel. Entering the sinusoidal steering zone (i.e., the sixth interval mentioned above), the steering wheel rotates in a sinusoidal manner. For example, the sinusoidal rotation can be expressed as follows:

[0235] Among them, δ max represents the peak steering wheel angle, T2-T1 represents the steering wheel rotation period (i.e., sinusoidal steering area), and t is the moment of sinusoidal rotation.

[0236] It can be seen that in the non-defensive emergency steering model in Figure 5, the parameters to be confirmed for constructing the non-defensive emergency steering model may include but are not limited to: the time point of risk occurrence, the duration of the straight-line period, the yaw moment of inertia, the steering free travel, the steering transmission ratio, the steering wheel angle peak, the steering wheel rotation cycle, and the like.

[0237] The third reaction model constructed through the embodiment of the present application can cover driving scenarios where real risk points appear, optimize the definition logic of real risk points, and more accurately characterize the vehicle's defenseless emergency steering behavior after the real risk points appear, thereby making a reasonable assessment of the autonomous driving capabilities and improving the safety of autonomous driving.

[0238] Case 4:

[0239] In the case where the first information includes the second risk point, fifth information is determined, and a fourth reaction model is constructed based on the first information and the fifth information.

[0240] When the first information includes a second risk point, it indicates that the road currently occupied by the first vehicle is a potential risk point. This can be used to provide a foreshadowing warning of a third risk point (i.e., a real risk point) that is reasonably foreseeable within a first time period after the current moment. In this case, the determined fifth information includes information about the seventh section in which the first vehicle travels in a straight line and information about the eighth section in which the vehicle turns.

[0241] Correspondingly, the fourth reaction model constructed based on the first information and the fifth information is used to characterize the defensive emergency steering behavior of the first vehicle in this scenario after the second risk point appears, that is, after the second risk point appears and before the third risk point appears, the first vehicle maintains a gliding state, that is, it drives in the current driving state. After the third risk point appears, the first vehicle continues to drive in a straight line in the seventh interval, and then starts to turn the vehicle in the eighth interval, first reaching the maximum turning angle in a sinusoidal form and then returning to the positive turning angle, and maintaining the direction after returning to the positive turning angle to continue driving in a straight line.

[0242] It can be understood that when the first information includes the second risk point, the fifth information required to construct the fourth reaction model may include but is not limited to: the duration of the first vehicle's straight-line driving in the seventh interval, the peak steering wheel angle of the first vehicle when turning the vehicle in the eighth interval, the steering wheel rotation cycle of the first vehicle, and so on.

[0243] The fourth reaction model constructed based on the fifth information in the embodiment of the present application can more accurately characterize the vehicle's defenseless emergency steering behavior after the potential risk point appears.

[0244] Optionally, the fourth reaction model constructed in this embodiment is similar to the non-defensive emergency steering model in Figure 5 above. The difference between the two is that the seventh interval in the fourth reaction model is different from the straight-line driving state after the first risk point appears in the non-defensive emergency steering model in Figure 5 above (i.e., the above-mentioned fifth interval). Specifically, the starting time point, ending time point, interval length, etc. of the interval are different. The intervals of the two are respectively suitable for defensive steering scenarios and stress steering scenarios; the eighth interval in the fourth reaction model is different from the sinusoidal steering area (i.e., the above-mentioned sixth interval) in the non-defensive emergency steering model in Figure 5 above. Specifically, the starting time point, ending time point, interval length, etc. of the interval are different. The intervals of the two are respectively suitable for defensive steering scenarios and stress steering scenarios.

[0245] The fourth reaction model constructed through the embodiment of the present application introduces the concept of potential risk points, which can cover driving scenarios where potential risk points appear, and can optimize the definition logic of real risk points, and more accurately characterize the vehicle's defensive emergency steering behavior after the potential risk points appear, thereby making a reasonable assessment of the autonomous driving capability and improving the safety of autonomous driving.

[0246] Based on the classification of the two actions and two stages of the standard operations of general human drivers by the reaction model in Figure 1 above, and the reaction model construction method in Figure 2 above, the embodiment of the present application correspondingly provides a parameter calibration method for the reaction model.

[0247] Please refer to Figure 6, which is a flow chart of a parameter calibration method for a reaction model provided in an embodiment of the present application. It is understandable that the steps in the embodiment of the present application can be regarded as a reasonable variation or supplement of the embodiment in Figure 2 above; or, it is understandable that the parameter calibration method for the reaction model in the embodiment of the present application can also be regarded as an embodiment that can be executed independently, and the present application does not limit this. The parameter calibration method for the reaction model in the embodiment of the present application is applied to the field of intelligent driving technology, such as the evaluation of autonomous driving capabilities. The parameter calibration method for the reaction model is specifically applied to the first vehicle, including but not limited to the following steps:

[0248] S601: Acquire sixth information.

[0249] S602: calibrate and obtain target parameters based on the sixth information.

[0250] It is understandable that the communication device used to obtain the sixth information and calibrate the target parameters based on the sixth information in the embodiment of the present application can be a device equipped with a processor / chip that can be used to execute computer-executable instructions, or a processor / chip that can be used to execute computer-executable instructions. The embodiment of the present application does not limit this. Optionally, the communication device can be a processor / chip in a smart car, or a simulated driving cabin and motion platform, which is used to execute the parameter calibration method of the reaction model in the embodiment of the present application to achieve coverage of more dangerous driving scenarios and more accurate characterization of the vehicle's behavior after the risk point appears, thereby making a reasonable assessment of the ability of autonomous driving and improving the safety of autonomous driving.

[0251] The sixth information mentioned above includes parameter information output by the reaction model.

[0252] The parameter information is used to characterize the driving behavior of a general human driver corresponding to the first vehicle. The reaction model is constructed based on the first information and is used to characterize the action behavior of the first vehicle after the first risk point or the second risk point occurs.

[0253] Specifically, the first information includes information of the first risk point or the second risk point.

[0254] The first risk point is a real risk point that appears on the road where the first vehicle is located at the first moment. The real risk point can be understood as an event that threatens the driving safety of the vehicle and occurs at the first moment.

[0255] The second risk point is a potential risk point that appears on the road where the first vehicle is located at the first moment and triggers the third risk point. The third risk point is a real risk point that appears on the road where the first vehicle is located within the first time period after the first moment. The potential risk point can be understood as an event that occurs at the first moment and may affect the vehicle driving safety, but has not yet posed an actual threat to the vehicle driving safety. The subsequent events triggered by it pose a threat to the vehicle driving safety.

[0256] It is understandable that the second risk point can be used to provide a foreshadowing warning for a reasonably foreseeable real risk point within a first time period after the first moment. The first time period is not a fixed value and can be adjusted according to different application scenarios.

[0257] Optionally, the reaction model used to output parameter information in the embodiment of the present application can be obtained by the reaction model construction method shown in Figure 2 above. For this, please refer to the description of the reaction model construction method shown in Figure 2 above, and the embodiment of the present application will not be repeated.

[0258] In the embodiment of the present application, the target parameters obtained based on the sixth information calibration are used to evaluate the autonomous driving performance of the first vehicle.

[0259] In an embodiment of the present application, target parameters are obtained by calibration based on the parameter information output by the reaction model, which are used to evaluate the autonomous driving performance of the first vehicle. Since the concept of potential risk points is introduced in the reaction model, more dangerous driving scenarios can be covered, and the definition logic of real risk points can be optimized to more accurately characterize the vehicle's action behavior after the risk points appear. Therefore, the target parameters obtained by calibration based on the parameter information output by the reaction model can make a reasonable assessment of the autonomous driving capability, improve the safety of autonomous driving, and also judge subsequent autonomous driving accidents.

[0260] From the reaction model construction method in Figure 2 above, it can be seen that the reaction models (braking model / steering model) constructed in different driving scenarios (real risk points / potential risk points) are different, for example:

[0261] In case 1 of the reaction model construction method in Figure 2 above, the first reaction model constructed based on the first information and the second information is used to characterize the defenseless emergency braking behavior of the first vehicle after the first risk point appears in a scenario where there is a first risk point (real risk point).

[0262] In the second case of the reaction model construction method in Figure 2 above, the second reaction model constructed based on the first information and the third information is used to characterize the defensive emergency braking behavior of the first vehicle after the second risk point appears in a scenario where there is a second risk point (potential risk point).

[0263] In the third case of the reaction model construction method in the above-mentioned Figure 2, the third reaction model constructed based on the first information and the fourth information is used to characterize the defenseless emergency steering behavior of the first vehicle after the first risk point appears in the scenario where the first risk point (real risk point) exists.

[0264] In case four of the reaction model construction method in Figure 2 above, the fourth reaction model constructed based on the first information and the fifth information is used to characterize the defensive emergency steering behavior of the first vehicle after the second risk point appears in a scenario where there is a second risk point (potential risk point).

[0265] In a possible embodiment, each reaction model (first reaction model, second reaction model, third reaction model, fourth reaction model) constructed by the model construction method in FIG. 2 can be combined to further illustrate step S602 in FIG. 6 “obtaining target parameters based on the sixth information calibration” according to different situations. The specific situations are as follows:

[0266] Case 1:

[0267] In the case where the parameter information included in the sixth information comes from the first reaction model or the second reaction model, the target parameters obtained by calibration based on the sixth information include but are not limited to any one or more of the following:

[0268] The maximum braking deceleration of the first vehicle, the braking effectiveness improvement time of the first vehicle, the braking effectiveness improvement rate of the first vehicle, and the decision response time of an average human driver corresponding to the first vehicle.

[0269] Furthermore, the calibration methods of various target parameters are described below.

[0270] (1) Maximum braking deceleration of the first vehicle:

[0271] First, the braking deceleration of the driver's vehicle, as output by the reaction model, is smoothed and filtered to remove a few outliers caused by collisions or system offsets. Next, a slice of data spanning a certain number of frames (e.g., 500 frames) before and after the risk point is selected for analysis. The average of N frames (e.g., 10 frames) with the highest braking deceleration after the occurrence of the first or second risk point is taken as the maximum braking deceleration. This more accurately characterizes the braking behavior of a typical human driver of the first vehicle after the occurrence of the first or second risk point.

[0272] (2) Braking efficiency improvement time of the first vehicle:

[0273] The braking start time is obtained by calibrating at least two frames of data with smaller braking deceleration in the parameter information output by the reaction model, and the braking saturation time is obtained by calibrating at least two frames of data with larger braking deceleration in the parameter information output by the reaction model. The difference between the braking saturation time and the braking start time is the braking efficiency improvement time of the first vehicle.

[0274] It is understood that a low braking deceleration may refer to a braking deceleration less than a preset threshold value. This preset threshold value is not a fixed value and can be adjusted according to different application scenarios. For example, it may be two frames of data with braking decelerations of 10% and 15% of the maximum braking deceleration. Similarly, a high braking deceleration may refer to a braking deceleration greater than a preset threshold value. This preset threshold value is not a fixed value and can be adjusted according to different application scenarios. For example, it may be two frames of data with braking decelerations of 90% and 85% of the maximum braking deceleration.

[0275] For example, the moment when the braking deceleration of the first frame reaches 10% of the maximum braking deceleration is recorded as T 10 The moment when the braking deceleration of the first frame reaches 15% of the maximum braking deceleration is recorded as T 15 , then the approximate braking start time T0 can be obtained as: T0=T10 -2×(T 15 -T 10 )

[0276] Similarly, find the moment when the braking deceleration of the first frame reaches 90% of the maximum braking deceleration and record it as T 90 , find the moment when the braking deceleration of the first frame reaches 85% of the maximum braking deceleration and record it as T 85 , then we can get the approximate braking saturation time T 100 For: T 100 =T 90 +2×(T 90 -T 85 )

[0277] Finally, T 100 Subtract T0 to obtain the braking efficiency improvement time of the first vehicle.

[0278] In the embodiment of the present application, since the brake pedal often produces slight disturbances during actual driving, it is impossible to directly find the moment when the braking deceleration of the first frame is not zero in the statistical code, otherwise it will be inaccurate due to the disturbance. To solve this problem, through the approximation method, it is assumed that the braking efficiency improvement rate is the same when braking starts and when it is about to be saturated. The braking saturation moment is reversed using the moment when the braking deceleration is larger, and the braking start moment is reversed using the moment when the braking deceleration is smaller. The robust and anti-interference braking efficiency improvement time can be obtained, so that the braking behavior of the general human driver corresponding to the first vehicle after the first risk point or the second risk point appears can be more accurately characterized.

[0279] (3) Braking efficiency improvement rate of the first vehicle:

[0280] The braking efficiency improvement rate of the first vehicle is obtained by calibrating at least one frame of data with a smaller braking deceleration and at least one frame of data with a larger braking deceleration in the parameter information output by the reaction model.

[0281] It is understood that "low braking deceleration" may refer to a braking deceleration less than a preset threshold value. This preset threshold value is not a fixed value and can be adjusted according to different application scenarios. For example, it may refer to at least one frame of data in which the braking deceleration is 10% of the maximum braking deceleration. Similarly, "high braking deceleration" may refer to a braking deceleration greater than a preset threshold value. This preset threshold value is not a fixed value and can be adjusted according to different application scenarios. For example, it may refer to at least one frame of data in which the braking deceleration is 90% of the maximum braking deceleration.

[0282] For example, if the maximum braking deceleration is d max , then the braking efficiency improvement rate of the first vehicle is calculated as follows:

[0283] In the embodiment of the present application, the braking curve is not a strictly straight line during the braking process. When braking starts and is about to be saturated, the braking deceleration will increase more slowly. Therefore, by selecting the braking deceleration change in the middle section of the entire braking process of the first vehicle, a more accurate braking efficiency improvement rate of the first vehicle can be obtained, thereby more accurately characterizing the braking behavior of the general human driver corresponding to the first vehicle after the first risk point or the second risk point appears.

[0284] (4) Decision response time of an average human driver corresponding to the first vehicle:

[0285] The braking start time is obtained by calibrating at least two frames of data with smaller braking deceleration in the parameter information output by the reaction model. The difference between the braking start time and the time when the actual risk point occurs is used as the decision response time of the average human driver corresponding to the first vehicle. This can more accurately characterize the braking behavior of the average human driver corresponding to the first vehicle after the first risk point or the second risk point occurs.

[0286] It can be understood that a smaller braking deceleration may refer to a braking deceleration that is less than a preset threshold value. The preset threshold value is not a fixed value and can be adjusted according to different application scenarios. For example, it may be at least two frames of data with a braking deceleration of 10% and 15% of the maximum braking deceleration.

[0287] Case 2:

[0288] In the case where the parameter information included in the sixth information comes from the third reaction model or the fourth reaction model, the target parameters obtained by calibration based on the sixth information include but are not limited to any one or more of the following:

[0289] a maximum steering wheel turning angle of the first vehicle, a steering wheel return time of the first vehicle, and a decision response time of an average human driver corresponding to the first vehicle.

[0290] Furthermore, the calibration methods of various target parameters are described below.

[0291] (1) Maximum steering wheel angle of the first vehicle:

[0292] First, the driver's steering wheel angle data from the parameter information output by the reaction model is smoothed and filtered to remove a few singular data points caused by collisions and system offsets. Then, a slice of data spanning a certain number of frames (e.g., 500 frames) before and after the risk point is selected for analysis. Within this slice, the M frames (e.g., 10 frames) with the largest steering wheel angles after the occurrence of the first or second risk point are averaged and used as the maximum steering wheel angle. This allows for a more accurate representation of the steering behavior of a typical human driver of the first vehicle after the occurrence of the first or second risk point.

[0293] (2) Time it takes for the steering wheel of the first vehicle to return to center:

[0294] The turning start time is obtained by calibrating at least two frames of data with smaller steering wheel angles in the parameter information output by the reaction model, and the first return time is obtained by calibrating at least two frames of data with larger steering wheel angles in the parameter information output by the reaction model. The difference between the first return time and the turning start time is the steering wheel return time of the first vehicle.

[0295] It is understood that a small steering wheel angle may refer to a steering wheel angle less than a certain preset threshold value, which is not a fixed value and can be adjusted according to different application scenarios. For example, it may be two frames of data with steering wheel angles of 20% and 15% of the maximum steering wheel angle. Similarly, a large steering wheel angle may refer to a steering wheel angle greater than a certain preset threshold value, which is not a fixed value and can be adjusted according to different application scenarios. For example, it may be two frames of data with steering wheel angles of 80% and 85% of the maximum steering wheel angle.

[0296] For example, the moment when the steering wheel angle reaches 20% of the maximum steering wheel angle in the first frame is found and recorded as t 20 The moment when the steering wheel angle in the first frame reaches 15% of the maximum steering wheel angle is recorded as t 15 , then the approximate turning start time t1 can be obtained as: t1 = t 15 -3×(t 10 -t 15 )

[0297] Similarly, find the moment when the steering wheel angle reaches 80% of the maximum steering wheel angle in the first frame and record it as t 80 , find the moment when the steering wheel angle in the first frame reaches 85% of the maximum steering wheel angle and record it as t 85 , then the approximate first steering return time t2 can be obtained as: t2 = t 85 +3×(t 85 -t 80 )

[0298] Finally, t1 is subtracted from t2 to obtain the time it takes for the steering wheel of the first vehicle to return to center.

[0299] In an embodiment of the present application, since the steering wheel often produces slight disturbances during actual driving, it is impossible to directly find the moment when the steering wheel angle is not zero in the first frame in the statistical code, otherwise it will be inaccurate due to the disturbance. To solve this problem, through the approximation method, it is assumed that the steering angle increase rate is the same at the beginning of steering and when the first steering return is about to begin. The moment at a larger steering angle is used to reverse the first steering return moment, and the moment at a smaller steering angle is used to reverse the steering start moment. The robust and anti-interference steering wheel return time can be obtained, so that the steering behavior of the general human driver corresponding to the first vehicle after the first risk point or the second risk point appears can be more accurately characterized.

[0300] (3) Decision response time of an average human driver corresponding to the first vehicle:

[0301] The turning start time is obtained by calibrating the data of at least two frames with smaller steering wheel angles in the parameter information. The difference between the turning start time and the time when the actual risk point occurs is used as the decision response time of the average human driver corresponding to the first vehicle. This can more accurately characterize the steering behavior of the average human driver corresponding to the first vehicle after the first risk point or the second risk point appears.

[0302] It can be understood that a smaller steering wheel angle may mean that the steering wheel angle is less than a preset threshold. The preset threshold is not a fixed value and can be adjusted according to different application scenarios. For example, it can be at least two frames of data where the steering wheel angle is 20% and 15% of the maximum steering wheel angle.

[0303] In an embodiment of the present application, target parameters are obtained by calibration based on the parameter information output by the reaction model, which are used to evaluate the autonomous driving performance of the first vehicle. Since the concept of potential risk points is introduced in the reaction model, more dangerous driving scenarios can be covered, and the definition logic of real risk points can be optimized to more accurately characterize the vehicle's action behavior after the risk points appear. Therefore, the target parameters obtained by calibration based on the parameter information output by the reaction model can make a reasonable assessment of the autonomous driving capability, improve the safety of autonomous driving, and also judge subsequent autonomous driving accidents.

[0304] In addition, in combination with the reaction model construction method shown in Figure 2 and the reaction model parameter calibration method shown in Figure 6, experiments can be designed for each driving scenario shown in Figures 4A, 4B and 4C based on a simulated driving cockpit and a motion platform. According to the reaction model parameter calibration method shown in Figure 6, the parameters of the reaction model corresponding to the average human driver in each driving scenario can be calibrated.

[0305] Specifically, for the scenario shown in FIG4A above where a small vehicle cuts in without turning on the turn signal, a series of experiments were designed to obtain the parameter results of the reaction model corresponding to the average human driver in this scenario, as shown in Tables 1 and 2 below:

[0306] Table 1

[0307] Table 2

[0308] Among them, the above Table 1 shows the parameter results of the braking model corresponding to the general human driver after the real risk point appears, and the above Table 2 shows the parameter results of the steering model corresponding to the general human driver after the real risk point appears.

[0309] Specifically, for the scenario shown in FIG4B above where a large vehicle cuts in without turning on the turn signal, a series of experiments were designed to obtain the parameter results of the reaction model corresponding to the average human driver in this scenario, as shown in Tables 3 and 4 below:

[0310] Table 3

[0311] Table 4

[0312] Among them, the above Table 3 shows the parameter results of the braking model corresponding to the general human driver after the potential risk point appears, and the above Table 4 shows the parameter results of the steering model corresponding to the general human driver after the potential risk point appears.

[0313] Specifically, for the preceding vehicle cut-out scenario shown in FIG4C , a series of experiments were designed to obtain the parameter results of the reaction model corresponding to the average human driver in this scenario, as shown in Tables 5 and 6 below:

[0314] Table 5

[0315] Table 6

[0316] Among them, the above Table 5 shows the parameter results of the braking model corresponding to the general human driver after the real risk point appears, and the above Table 6 shows the parameter results of the steering model corresponding to the general human driver after the real risk point appears.

[0317] It should be understood that the above Tables 1 to 6 are merely several sets of possible test data to exemplify the effects of the reaction model construction method shown in FIG2 and the parameter calibration method of the reaction model shown in FIG6 , and should not be used to limit the embodiments of the present application. The test data obtained based on the supplement or reasonable deformation of the above driving scenarios and design experiments to support the corresponding effects of the above methods are all within the scope of protection of the embodiments of the present application.

[0318] The above describes in detail the methods of the embodiments of the present application. The following provides an apparatus for implementing any method in the embodiments of the present application. For example, an apparatus is provided that includes units (or means) for implementing each step performed by the device in any of the above methods.

[0319] Please refer to FIG. 7 , which is a schematic structural diagram of a communication device provided in an embodiment of the present application.

[0320] As shown in Figure 7, the communication device 70 may include a communication unit 701 and a processing unit 702. The communication unit 701 and the processing unit 702 may be software, hardware, or a combination of software and hardware.

[0321] The communication unit 701 can implement a sending function and / or a receiving function, and can also be described as a transceiver unit. The communication unit 701 can also be a unit that integrates an acquisition unit and a transmission unit, wherein the acquisition unit is used to implement the receiving function and the transmission unit is used to implement the transmission function. Optionally, the communication unit 701 can be used to receive information sent by other devices, and can also be used to send information to other devices.

[0322] In one possible design, the communication device 70 may correspond to the communication device in the method embodiment shown in FIG. 2 , for example, the communication device 70 may be the communication device in the method embodiment shown in FIG. 2 , or may be a chip in the communication device. The communication device 70 may include a unit for executing the operations performed by the communication device in the method embodiment shown in FIG. 2 , and each unit in the communication device 70 is for implementing the operations performed by the communication device in the method embodiment shown in FIG. 2 . The description of each unit is as follows:

[0323] Processing unit 702 is configured to obtain first information, where the first information includes information about a first risk point or a second risk point, where the first risk point is a real risk point on the road where the first vehicle is currently located; the second risk point is a potential risk point on the road where the first vehicle is currently located that may trigger a third risk point; and the third risk point is a real risk point on the road where the first vehicle is currently located that may occur within a first time period after the current time.

[0324] The processing unit 702 is further configured to construct a reaction model based on the first information, where the reaction model is configured to characterize an action behavior of the first vehicle after the first risk point or the second risk point occurs.

[0325] In one possible implementation, the device further includes:

[0326] The communication unit 701 is configured to receive the first information;

[0327] The processing unit 702 is specifically configured to obtain the first information through the communication unit 701 .

[0328] The method executed by the processing unit 702 and the communication unit 701 may refer to the method corresponding to FIG2 , which will not be described in detail here.

[0329] Regarding the technical effects brought about by this design and any possible implementation method, please refer to the introduction of the technical effects of the corresponding method in Figure 2 above.

[0330] In another possible design of the communication device 70 shown in FIG7 , the communication device 70 may correspond to the communication device in the method embodiment shown in FIG6 . For example, the communication device 70 may be the communication device in the method embodiment shown in FIG6 , or may be a chip in the communication device. The communication device 70 may include a unit for performing the operations performed by the communication device in the method embodiment shown in FIG6 , and each unit in the communication device 70 is respectively for implementing the operations performed by the communication device in the method embodiment shown in FIG6 . The description of each unit is as follows:

[0331] Processing unit 702 is configured to obtain sixth information, the sixth information including parameter information output by a reaction model, the parameter information being used to characterize driving behavior of a typical human driver corresponding to the first vehicle; wherein the reaction model is constructed based on the first information, the first information including information about a first risk point or a second risk point, the first risk point being an actual risk point appearing on the road where the first vehicle is located at a first moment, the second risk point being a potential risk point appearing on the road where the first vehicle is located at a first moment that triggers a third risk point, the third risk point being an actual risk point appearing on the road where the first vehicle is located within a first time period after the first moment, and the reaction model being used to characterize the first vehicle's action after the occurrence of the first risk point or the second risk point;

[0332] The processing unit 702 is further configured to obtain target parameters based on the sixth information calibration, and the target parameters are used to evaluate the autonomous driving performance of the first vehicle.

[0333] In one possible implementation, the device further includes:

[0334] The communication unit 701 is configured to receive the sixth information;

[0335] The processing unit 702 is specifically configured to obtain the sixth information through the communication unit 701 .

[0336] The method executed by the processing unit 702 and the communication unit 701 may refer to the method corresponding to FIG6 , which will not be described again here.

[0337] Regarding the technical effects brought about by this design and any possible implementation method, please refer to the introduction of the technical effects of the corresponding method in Figure 6 above.

[0338] Optionally, in the communication device described in any of the foregoing designs and any of the possible implementations:

[0339] In one implementation, the communication apparatus is a communication device. When the communication apparatus is a communication device, the communication unit may be a transceiver or an input / output interface; and the processing unit may be at least one processor. Alternatively, the transceiver may be a transceiver circuit. Alternatively, the input / output interface may be an input / output circuit.

[0340] In another implementation, the communication device is a chip (system) or circuit used in a communication device. When the communication device is a chip (system) or circuit used in a communication device, the communication unit may be a communication interface (input / output interface), interface circuit, output circuit, input circuit, pin, or related circuit on the chip (system) or circuit; and the processing unit may be at least one processor, processing circuit, or logic circuit.

[0341] According to an embodiment of the present application, the various units in the device shown in Figure 7 can be separately or all merged into one or several other units to constitute, or one (some) of the units can also be split into multiple smaller units in function to constitute, which can achieve the same operation without affecting the realization of the technical effects of the embodiments of the present application. The above-mentioned units are divided based on logical functions. In practical applications, the functions of a unit can also be implemented by multiple units, or the functions of multiple units can be implemented by one unit. In other embodiments of the present application, other units can also be included based on electronic equipment. In practical applications, these functions can also be implemented with the assistance of other units, and can be implemented by the collaboration of multiple units.

[0342] It should be noted that the implementation of each unit may also refer to the corresponding description of the method embodiments shown in FIG. 2 and FIG. 6 .

[0343] In the communication device 70 described in Figure 7, by introducing the concept of potential risk points, more dangerous driving scenarios can be covered, and the definition logic of real risk points can be optimized, so as to more accurately characterize the vehicle's action behavior after the risk points appear, thereby making a reasonable assessment of the autonomous driving capabilities and improving the safety of autonomous driving.

[0344] Please refer to FIG8 , which is a schematic structural diagram of a communication device provided in an embodiment of the present application.

[0345] It should be understood that the communication device 80 shown in Figure 8 is only an example. The communication device of the embodiment of the present application may also include other components, or include components with similar functions to the components in Figure 8, or not include all the components in Figure 8.

[0346] The communication device 80 includes a communication interface 801 and at least one processor 802 .

[0347] The communication interface 801 is used to send and receive signals, and the at least one processor 802 executes program instructions, so that the communication device 80 implements the corresponding process of the method executed by the corresponding device in the above method embodiment.

[0348] In one possible design, the communication device 80 may correspond to the communication device in the method embodiment shown in FIG. 2 . For example, the communication device 80 may be the communication device in the method embodiment shown in FIG. 2 , or may be a chip in the communication device. The communication device 80 may include components for executing the operations performed by the communication device in the method embodiment described above, and each component in the communication device 80 is for implementing the operations performed by the communication device in the method embodiment described above. Specifically, the operations may be as follows:

[0349] Processor 802 is configured to obtain first information, where the first information includes information about a first risk point or a second risk point, where the first risk point is a real risk point on the road where the first vehicle is currently located; the second risk point is a potential risk point on the road where the first vehicle is currently located that may trigger a third risk point; and the third risk point is a real risk point on the road where the first vehicle is currently located that may occur within a first time period after the current moment.

[0350] The processor 802 is further configured to construct a reaction model based on the first information, where the reaction model is configured to characterize an action behavior of the first vehicle after the first risk point or the second risk point occurs.

[0351] In one possible implementation, the device further includes:

[0352] Communication interface 801, used for receiving the first information;

[0353] The processor 802 is specifically configured to obtain the first information through the communication interface 801 .

[0354] In another possible design, the communication device 80 may correspond to the communication device in the method embodiment shown in FIG. 6 . For example, the communication device 80 may be the communication device in the method embodiment shown in FIG. 6 , or may be a chip in the communication device. The communication device 80 may include components for executing the operations performed by the communication device in the method embodiment described above, and each component in the communication device 80 is for implementing the operations performed by the communication device in the method embodiment described above. Specifically, the operation may be as follows:

[0355] Processor 802 is configured to obtain sixth information, the sixth information including parameter information output by a reaction model, the parameter information being used to characterize driving behavior of a typical human driver corresponding to the first vehicle; wherein the reaction model is constructed based on the first information, the first information including information about a first risk point or a second risk point, the first risk point being an actual risk point appearing on the road where the first vehicle is located at a first moment, the second risk point being a potential risk point appearing on the road where the first vehicle is located at a first moment that triggers a third risk point, the third risk point being an actual risk point appearing on the road where the first vehicle is located within a first time period after the first moment, and the reaction model being used to characterize the first vehicle's action after the occurrence of the first risk point or the second risk point;

[0356] The processor 802 is further configured to obtain target parameters based on the sixth information calibration, and the target parameters are used to evaluate the autonomous driving performance of the first vehicle.

[0357] In one possible implementation, the device further includes:

[0358] Communication interface 801, used to receive the sixth information;

[0359] The processor 802 is specifically configured to obtain the sixth information through the communication interface 801 .

[0360] In the communication device 80 described in Figure 8, by introducing the concept of potential risk points, more dangerous driving scenarios can be covered, and the definition logic of real risk points can be optimized, so as to more accurately characterize the vehicle's action behavior after the risk points appear, thereby making a reasonable assessment of the autonomous driving capabilities and improving the safety of autonomous driving.

[0361] In the case where the communication device may be a chip or a chip system, reference may be made to the schematic structural diagram of the chip shown in FIG9 .

[0362] As shown in Figure 9, chip 90 includes a processor 901 and an interface 902. There may be one or more processors 901, and there may be multiple interfaces 902. It should be noted that the functions corresponding to processor 901 and interface 902 can be implemented through hardware design, software design, or a combination of hardware and software, without limitation.

[0363] Optionally, the chip 90 may further include a memory 903 , which is used to store necessary program instructions and data.

[0364] In this application, processor 901 may be configured to call a program for implementing the method provided in one or more embodiments of this application in a communication device from memory 903 and execute the instructions contained in the program. Interface 902 may be configured to output the execution results of processor 901. In this application, interface 902 may be specifically configured to output various messages or information from processor 901.

[0365] Regarding the methods provided by one or more embodiments of the present application, reference may be made to the embodiments shown in FIG. 2 and FIG. 6 , which will not be described in detail here.

[0366] The processor in the embodiments of the present application may be a central processing unit (CPU), and the processor may also be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field programmable gate arrays (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor, etc.

[0367] The memory in the embodiments of the present application is used to provide storage space, in which data such as an operating system and computer programs can be stored. The memory includes, but is not limited to, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM), or compact disc read-only memory (CD-ROM).

[0368] According to the method provided in the embodiment of the present application, the embodiment of the present application also provides a computer-readable storage medium, in which a computer program is stored. When the computer program runs on one or more processors, the method shown in Figures 2 and 6 can be implemented.

[0369] According to the method provided in the embodiment of the present application, the embodiment of the present application also provides a computer program product, which includes a computer program. When the computer program runs on a processor, it can implement the method shown in Figures 2 and 6 above.

[0370] An embodiment of the present application also provides a vehicle side, which includes at least one communication device 70 or communication device 80 or chip 90 as described above, for executing the steps executed by the corresponding device in any of the embodiments of Figures 2 and 6 above.

[0371] An embodiment of the present application further provides a processing device, including a processor and an interface; the processor is used to execute the method in any of the above method embodiments.

[0372] It should be understood that the above-mentioned processing device can be a chip. For example, the processing device can be a field programmable gate array (FPGA), a general-purpose processor, a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic device, a discrete gate or transistor logic device, a discrete hardware component, a system on chip (SoC), a central processing unit (CPU), a network processor (NP), a digital signal processing circuit (DSP), a microcontroller unit (MCU), a programmable logic device (PLD) or other integrated chip. The various methods, steps and logic block diagrams disclosed in the embodiments of the present application can be implemented or executed. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc. The steps of the method disclosed in conjunction with the embodiments of the present application can be directly embodied as being executed by a hardware decoding processor, or can be executed by a combination of hardware and software modules in the decoding processor. The software module can be located in a storage medium well-known in the art, such as random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, etc. The storage medium is located in the memory, and the processor reads the information in the memory and, in conjunction with its hardware, completes the steps of the above method.

[0373] It is understood that the memory in the embodiments of the present application may be a volatile memory or a non-volatile memory, or may include both volatile and non-volatile memories. Among them, the non-volatile memory may be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), or a flash memory. The volatile memory may be a random access memory (RAM), which is used as an external cache. By way of example and not limitation, many forms of RAM are available, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), and direct RAM bus RAM (DR RAM). It should be noted that the memory of the systems and methods described herein is intended to include, but is not limited to, these and any other suitable types of memory.

[0374] In the above embodiments, all or part of the embodiments may be implemented by software, hardware, firmware, or any combination thereof. When implemented using software, all or part of the embodiments may be implemented in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of the present application are generated. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions may be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions may be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via a wired (e.g., coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) method. The computer-readable storage medium may be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more available media integrated therein. The available medium may be a magnetic medium (eg, a floppy disk, a hard disk, a magnetic tape), an optical medium (eg, a high-density digital video disc (DVD)), or a semiconductor medium (eg, a solid state disc (SSD)).

[0375] The units in the above-mentioned various apparatus embodiments completely correspond to the electronic devices in the method embodiments, and the corresponding modules or units perform the corresponding steps. For example, the communication unit (transceiver) performs the receiving or sending steps in the method embodiments, and other steps except sending and receiving can be performed by the processing unit (processor). The functions of the specific units can be referred to the corresponding method embodiments. Among them, there can be one or more processors.

[0376] It is understood that in the embodiments of the present application, the electronic device can perform some or all of the steps in the embodiments of the present application. These steps or operations are merely examples, and the embodiments of the present application can also perform other operations or variations of various operations. In addition, the various steps can be performed in a different order than those presented in the embodiments of the present application, and it is possible that not all operations in the embodiments of the present application need to be performed.

[0377] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0378] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0379] In the several embodiments provided in this application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are merely schematic. For example, the division of the units is merely a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0380] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0381] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0382] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory ROM, a random access memory RAM, a magnetic disk or an optical disk.

[0383] The above is only a specific implementation method of the present application, but the scope of protection of the present application is not limited thereto. Any technician familiar with this technical field can easily think of changes or replacements within the technical scope disclosed in this application, which should be covered by the scope of protection of the present application.

Claims

1. A reaction model construction method, applied to a first vehicle, characterized in that: include: Acquire first information, where the first information includes information of a first risk point or a second risk point, where the first risk point is a real risk point that appears on the road where the first vehicle is located at a current moment, and the second risk point is a potential risk point that appears on the road where the first vehicle is located at a current moment and triggers a third risk point, and the third risk point is a real risk point that appears on the road where the first vehicle is located within a first time period after the current moment; A reaction model is constructed based on the first information, where the reaction model is used to characterize the action behavior of the first vehicle after the first risk point or the second risk point occurs.

2. The method according to claim 1, characterized in that The second risk point includes any one or more of the following: Road sections with deceleration warnings, road sections with attention signs, road sections with static blind spots, vehicles in adjacent lanes braking, adjacent vehicles being large vehicles, and warning signals for adjacent vehicles.

3. The method according to claim 1 or 2, characterized in that: The action behavior includes at least one of a vehicle braking behavior or a vehicle steering behavior.

4. The method according to any one of claims 1 to 3, characterized in that The first risk point is that the offset between the second vehicle and the center line of the lane where the second vehicle is located is greater than a first threshold; wherein, the second vehicle is a vehicle that cuts in front of the first vehicle without turning on the turn signal, and the model of the second vehicle is not larger than the model of the first vehicle.

5. The method according to any one of claims 1 to 3, characterized in that The second risk point is that the model of the third vehicle is larger than that of the first vehicle, and the third risk point is that the third vehicle is laterally offset; wherein, the third vehicle is adjacent to the first vehicle, and the third vehicle is a vehicle that cuts in front of the first vehicle without turning on the turn signal.

6. The method according to any one of claims 1 to 3, characterized in that The first risk point is that the lateral body of the fourth vehicle appearing in the field of view of the first vehicle is greater than a second threshold; wherein the fourth vehicle, the first vehicle and the fifth vehicle are located in the same lane, and the fourth vehicle is located in front of the fifth vehicle, the fifth vehicle is located in front of the first vehicle, and the fifth vehicle is a vehicle that cuts out of the current lane.

7. The method according to any one of claims 1 to 6, characterized in that The step of constructing a reaction model based on the first information includes: In a case where the first information includes the first risk point, second information is determined, where the second information includes information of a first interval and a second interval, the first interval is an interval in which the first vehicle travels at a constant speed, the second interval is an interval in which the first vehicle brakes, and the second interval is located after the first interval; A first reaction model is constructed based on the first information and the second information. The first reaction model is used to characterize that after the first risk point occurs, the first vehicle travels at a constant speed in the first interval and brakes until it stops in the second interval.

8. The method according to claim 7, characterized in that The second information also includes any one or more of the following: The duration of the constant speed travel of the first vehicle in the first interval, the start time point of vehicle braking of the first vehicle in the second interval, the braking rate of the first vehicle, and the maximum braking deceleration of the first vehicle.

9. The method according to any one of claims 1 to 6, characterized in that The step of constructing a reaction model based on the first information includes: In a case where the first information includes the second risk point, determining third information, the third information including information of a third interval and a fourth interval, the third interval being an interval in which the first vehicle travels at a constant speed, the fourth interval being an interval in which the first vehicle performs vehicle braking, and the fourth interval being located after the third interval; Based on the first information and the third information, a second reaction model is constructed, and the second reaction model is used to characterize that after the second risk point occurs and before the third risk point occurs, the first vehicle travels in the current driving state, and after the third risk point occurs, the first vehicle travels at a constant speed in the third interval and brakes the vehicle until it stops in the fourth interval.

10. The method according to claim 9, characterized in that The third information also includes any one or more of the following: The duration of the constant speed driving of the first vehicle in the third interval, the starting time point of vehicle braking of the first vehicle in the fourth interval, the braking rate of the first vehicle, and the maximum braking deceleration of the first vehicle.

11. The method according to any one of claims 1 to 6, characterized in that The step of constructing a reaction model based on the first information includes: In a case where the first information includes the first risk point, determining fourth information, the fourth information including information of a fifth interval and a sixth interval, the fifth interval being an interval in which the first vehicle travels in a straight line, the sixth interval being an interval in which the first vehicle turns, and the sixth interval being located after the fifth interval; A third reaction model is constructed based on the first information and the fourth information. The third reaction model is used to characterize that after the first risk point occurs, the first vehicle travels in a straight line in the fifth interval and turns the vehicle in the sixth interval until returning to the normal direction.

12. The method according to claim 11, characterized in that The fourth information also includes any one or more of the following: The duration of the straight-line driving of the first vehicle in the fifth interval, the peak value of the steering wheel angle of the first vehicle when turning the vehicle in the sixth interval, and the steering wheel rotation cycle of the first vehicle.

13. The method according to any one of claims 1 to 6, characterized in that The step of constructing a reaction model based on the first information includes: In a case where the first information includes the second risk point, fifth information is determined, the fifth information including information of a seventh interval and an eighth interval, the seventh interval being an interval in which the first vehicle travels in a straight line, the eighth interval being an interval in which the first vehicle turns, and the eighth interval being located after the seventh interval; Based on the first information and the fifth information, a fourth reaction model is constructed, and the fourth reaction model is used to characterize that after the second risk point occurs and before the third risk point occurs, the first vehicle travels in the current driving state, and after the third risk point occurs, the first vehicle travels in a straight line within the seventh interval and turns the vehicle within the eighth interval until it returns to the normal direction.

14. The method according to claim 13, characterized in that The fifth information also includes any one or more of the following: The duration of the straight-line driving of the first vehicle in the seventh interval, the peak steering wheel angle of the first vehicle when the first vehicle turns in the eighth interval, and the steering wheel rotation cycle of the first vehicle.

15. A method for calibrating parameters of a reaction model, applied to a first vehicle, characterized in that: include: Acquire sixth information, the sixth information including parameter information output by a reaction model, the parameter information being used to characterize the driving behavior of a general human driver corresponding to the first vehicle; wherein the reaction model is constructed from the first information, the first information including information of a first risk point or a second risk point, the first risk point being a real risk point appearing on the road where the first vehicle is located at a first moment, the second risk point being a potential risk point appearing on the road where the first vehicle is located at a first moment that triggers a third risk point, the third risk point being a real risk point appearing on the road where the first vehicle is located within a first time period after the first moment, and the reaction model being used to characterize the action behavior of the first vehicle after the first risk point or the second risk point appears; A target parameter is obtained based on the sixth information calibration, and the target parameter is used to evaluate the autonomous driving performance of the first vehicle.

16. The method according to claim 15, characterized in that In the case where the reaction model includes a first reaction model or a second reaction model, the target parameter includes any one or more of the following: The maximum braking deceleration of the first vehicle, the braking effectiveness improvement time of the first vehicle, the braking effectiveness improvement rate of the first vehicle, and the decision response time of a general human driver corresponding to the first vehicle; Among them, the first reaction model is used to characterize the braking behavior of the first vehicle after the first risk point appears, and the second reaction model is used to characterize the braking behavior of the first vehicle after the second risk point appears.

17. The method according to claim 16, characterized in that The maximum braking deceleration of the first vehicle is obtained by calibrating N frames of data with larger braking deceleration in the parameter information after the first risk point or the second risk point appears, where N is an integer greater than 0.

18. The method according to claim 16, characterized in that The braking efficiency improvement time of the first vehicle is the difference between the braking saturation moment and the braking start moment, the braking start moment is calibrated by at least two frames of data with a smaller braking deceleration in the parameter information, and the braking saturation moment is calibrated by at least two frames of data with a larger braking deceleration in the parameter information.

19. The method according to claim 16, characterized in that The braking efficiency improvement rate of the first vehicle is calibrated by at least one frame of data with a smaller braking deceleration and at least one frame of data with a larger braking deceleration in the parameter information.

20. The method according to claim 16, characterized in that The decision response time of the general human driver corresponding to the first vehicle is the difference between the braking start time and the time when the real risk point appears, and the braking start time is calibrated by at least two frames of data with smaller braking deceleration in the parameter information.

21. The method according to claim 15, characterized in that In the case where the reaction model includes the third reaction model or the fourth reaction model, the parameter information includes any one or more of the following: a maximum steering angle of the first vehicle, a steering wheel return time of the first vehicle, and a decision response time of a general human driver corresponding to the first vehicle; Among them, the third reaction model is used to characterize the steering behavior of the first vehicle after the first risk point appears, and the fourth reaction model is used to characterize the steering behavior of the first vehicle after the second risk point appears.

22. The method according to claim 21, characterized in that The maximum steering wheel angle of the first vehicle is obtained by calibrating M frames of data with a larger steering wheel angle in the parameter information after the first risk point or the second risk point appears, where M is an integer greater than 0.

23. The method according to claim 21, characterized in that The steering wheel return time of the first vehicle is the difference between the first return time and the steering start time, the steering start time is calibrated by at least two frames of data with smaller steering wheel angles in the parameter information, and the first return time is calibrated by at least two frames of data with larger steering wheel angles in the parameter information.

24. The method according to claim 21, characterized in that The decision response time of the general human driver corresponding to the first vehicle is the difference between the turning start time and the time when the real risk point appears, and the turning start time is calibrated by at least two frames of data with smaller steering wheel angles in the parameter information.

25. A communication device, characterized in that: The method comprises a module or a unit for executing the method as claimed in any one of claims 1 to 14 or claims 15 to 24.

26. A communication device, characterized in that: include: processor; When the processor calls the computer program or instruction in the memory, the method according to any one of claims 1 to 14 is executed, or the method according to any one of claims 15 to 24 is executed.

27. A communication device, characterized in that: comprising a logic circuit and an interface, wherein the logic circuit and the interface are coupled; The interface is used to input data to be processed, the logic circuit processes the data to be processed according to the method as described in any one of claims 1 to 14 or claims 15 to 24 to obtain processed data, and the interface is used to output the processed data.

28. A computer-readable storage medium, characterized in that: include: The computer-readable storage medium is used to store instructions or computer programs; when the instructions or the computer program are executed, the method according to any one of claims 1 to 14 is implemented, or the method according to any one of claims 15 to 24 is implemented.

29. A computer program product, characterized in that include: instructions or computer programs; When the instructions or the computer program are executed, the method according to any one of claims 1 to 14 is executed, or the method according to any one of claims 15 to 24 is executed.

30. A vehicle end, characterized in that: Comprising the communication device according to claim 25, or the communication device according to claim 26, or the communication device according to claim 27.

Citation Information

Patent Citations

  • Reaction model construction method, parameter calibration method of reaction model and related device

    CN119066824A

  • Method and device for determining driving risk and storage medium

    CN111444468A

  • Vehicle driving risk early warning method and device, equipment and storage medium

    CN111489588A

  • Method and device for determining vehicle driving risk, medium and electronic device

    CN111710187A

  • Driving risk prediction method and device, electronic equipment and storage medium

    CN112232581A

Cited By

  • Cost function construction method, computer readable storage medium and program product

    CN120256793A

  • Auxiliary driving optimization method based on expected function safety risk assessment

    CN121734377A