Intelligent driving risk early-warning method and apparatus, electronic device, medium, and product

Abstract

An intelligent driving risk early-warning method and apparatus, an electronic device, and a storage medium. The intelligent driving risk early-warning method is applied to a first vehicle, and comprises the following steps: when it is determined that the first vehicle is in an intelligent driving state, acquiring vehicle traveling information (S10); when it is determined that the vehicle traveling information satisfies a preset potential risk condition, performing risk data detection (S20); and when risk evaluation is performed by means of a preset risk evaluation model on the basis of the risk data to determine that the first vehicle has an intelligent driving risk, outputting intelligent driving risk early-warning information (S30).

Description

Intelligent driving risk warning methods, devices, electronic equipment, media and products

[0001] Related applications

[0002] This application claims priority to Chinese patent application No. 202410507126.1, filed on April 25, 2024, the entire contents of which are incorporated herein by reference. Technical Field

[0003] This application relates to the field of vehicle technology, and in particular to an intelligent driving risk warning method, device, electronic device, storage medium and product. Background Technology

[0004] With the rapid development of intelligent driving technology, it is being applied to an increasing number of vehicles, bringing renewed attention to its safety. Currently, user manuals are primarily used to instruct drivers on how to safely use intelligent driving functions, improving their understanding and effectively preventing accidents caused by human error. However, intelligent driving systems themselves can also make decision-making errors due to inaccuracies or inaccuracies, leading to incorrect actions or loss of control. Therefore, the system's limitations still present significant safety risks. Summary of the Invention

[0005] The main purpose of this application is to provide a method, device, electronic device, storage medium and product for early warning of risks in intelligent driving, which aims to solve the technical problem of high safety risks caused by system limitations in the process of intelligent driving in related technologies.

[0006] To achieve the above objectives, this application provides an intelligent driving risk warning method, which is applied to a first vehicle and includes the following steps:

[0007] If it is determined that the first vehicle is in intelligent driving mode, obtain vehicle driving information;

[0008] If the vehicle driving information is determined to meet the preset potential risk conditions, risk data detection is performed.

[0009] If a risk assessment is conducted based on the risk data using a preset risk assessment model, and it is determined that the first vehicle has a risk of intelligent driving, an intelligent driving risk warning message is output.

[0010] This application also provides an intelligent driving risk warning device, which is applied to a first vehicle and includes:

[0011] The acquisition module is used to acquire vehicle driving information when it is determined that the first vehicle is in an intelligent driving state;

[0012] The detection module is used to perform risk data detection when it is determined that the vehicle driving information meets preset potential risk conditions;

[0013] The output module is used to output intelligent driving risk warning information when the first vehicle is determined to have intelligent driving risks by performing a risk assessment based on the risk data using a preset risk assessment model.

[0014] This application also provides an electronic device, which is a physical device, comprising: a memory, a processor, and a program of the intelligent driving risk warning method stored in the memory and executable on the processor. When the program of the intelligent driving risk warning method is executed by the processor, it can implement the steps of the intelligent driving risk warning method as described above.

[0015] This application also provides a medium, which is a computer-readable storage medium, on which a program for implementing an intelligent driving risk warning method is stored. When the program for the intelligent driving risk warning method is executed by a processor, it implements the steps of the intelligent driving risk warning method as described above.

[0016] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the intelligent driving risk warning method described above.

[0017] This application provides a method, device, electronic device, storage medium, and product for intelligent driving risk warning. The intelligent driving risk warning method is applied to a first vehicle with intelligent driving capabilities. First, by determining that the first vehicle is in an intelligent driving state, vehicle driving information is acquired, achieving monitoring of the first vehicle in intelligent driving mode. Then, by determining that the vehicle driving information meets preset potential risk conditions, risk data detection is performed, achieving risk monitoring when the first vehicle has potential risks. Finally, by performing a risk assessment based on the risk data using a preset risk assessment model, if it is determined that the first vehicle has intelligent driving risks, intelligent driving risk warning information is output, achieving timely risk warning. Thus, if intelligent driving decisions violate the driver's intentions or environmental safety intentions due to system limitations, the human driver can promptly identify and intervene to control the vehicle after receiving the intelligent driving risk warning information, achieving the purpose of risk avoidance. Therefore, it overcomes the technical defects of intelligent driving systems themselves, which may make decision-making errors due to accuracy or errors, resulting in incorrect actions or loss of control, and thus still have high safety risks due to system limitations. This method compensates for the safety loopholes in intelligent driving system limitations and improves the safety of intelligent driving processes. Attached Figure Description

[0018] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0019] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 is a flowchart illustrating the first embodiment of the intelligent driving risk warning method of this application;

[0021] Figure 2 is a flowchart illustrating an embodiment of the intelligent driving risk warning method of this application;

[0022] Figure 3 is a flowchart illustrating the second embodiment of the intelligent driving risk warning method of this application;

[0023] Figure 4 is a schematic diagram of an embodiment of the intelligent driving risk warning method device in this application;

[0024] Figure 5 is a schematic diagram of the hardware operating environment involved in the intelligent driving risk warning method in the embodiments of this application;

[0025] Figure 6 is a flowchart illustrating an embodiment of the intelligent driving risk warning method of this application;

[0026] Figure 7 is a flowchart illustrating an embodiment of the intelligent driving risk warning method of this application.

[0027] The purpose, features and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0028] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0029] Example 1

[0030] This application provides an intelligent driving risk warning method. In a first embodiment of the intelligent driving risk warning method of this application, referring to FIG1, the intelligent driving risk warning method is applied to a first vehicle and includes the following steps:

[0031] Step S10: If it is determined that the first vehicle is in intelligent driving state, obtain vehicle driving information;

[0032] The subject executing the method in this embodiment can be an intelligent driving risk warning device, an intelligent driving risk warning terminal device, or a server. This embodiment takes an intelligent driving risk warning device as an example. The intelligent driving risk warning device can be integrated into terminal devices such as vehicles with data processing functions, in-vehicle terminals, vehicle controllers, smartphones, and tablets.

[0033] In this embodiment, the first vehicle has intelligent driving functionality. In one embodiment, the intelligent driving functionality can be ADAS (Advanced Driving Assistance System) functionality. Intelligent driving state refers to normal intelligent driving cruise conditions, such as ACC (Adaptive Cruise Control) and HWA (Highway Assist) modes. The intelligent driving risk warning method is used to provide risk warnings to the first vehicle during intelligent driving under intelligent driving control. In one embodiment, the intelligent driving risk warning device can be deployed on the first vehicle; in another embodiment, the intelligent driving risk warning device can be deployed on the server. The specific deployment method can be determined according to the actual situation, and this embodiment does not impose any restrictions.

[0034] In one embodiment, the first vehicle further includes an intelligent driving system, which includes at least one sensing module, such as radar or a camera. Each sensing module independently senses objects in the external environment and generates object information, so that the intelligent driving system can make intelligent driving decisions based on the object information.

[0035] Risk data refers to data that may lead to risks for the first vehicle in intelligent driving mode. It can include one or more of the following: vehicle status information, environmental perception information, and intelligent driving verification information. Vehicle status information refers to the status information of the first vehicle, including intelligent driving speed and vehicle speed. For example, if there are abnormalities in the intelligent driving speed sensor or vehicle speed sensor, or if there is a delay in data transmission, intelligent driving decisions such as excessive speed or excessive angle can easily cause safety accidents. Environmental perception information refers to information about the environment in which the first vehicle is located, including road information, lane reference lines, and object information. Objects include pedestrians, traffic facilities, and other vehicles besides the first vehicle. Object information includes obstacle positions and object speeds. When intelligent driving decisions violate environmental safety intentions, such as when there is an intersection nearby or the vehicle is making a sharp turn, inaccurate timing, excessively large or small angles in intelligent driving can easily lead to collisions. Intelligent driving verification information refers to relevant information used to verify the accuracy of intelligent driving decisions. This includes the confidence level of each perception module's independent perception of its environment, the deviation between the perception capabilities of different perception modules, and sensor accuracy. For example, a vehicle equipped with radar and cameras may need to turn right at an intersection, but there may be pedestrians crossing the road. Sometimes, these pedestrians may be obscured by other vehicles or pedestrians, making them undetectable. If the intelligent driving system controls the vehicle based on perception data that fails to detect the pedestrians, it could lead to a safety accident. Similarly, delays in camera data transmission or image recognition can cause significant differences between the pedestrian's position perceived by the camera and that perceived by the radar. In such cases, if the intelligent driving system controls the vehicle based on the pedestrian's position perceived by the camera, it could also lead to a safety accident. The perception accuracy of the perception modules is a crucial factor affecting intelligent driving decisions. However, environmental perception technology, sensor accuracy, and data transmission delays can all affect the accuracy of the perception modules' perception of the external environment, thus impacting the accuracy of intelligent driving decisions.

[0036] Vehicle driving information is used to characterize the current driving state of the first vehicle, including vehicle speed, acceleration, steering speed, fault information, etc. When the first vehicle's driving state is relatively smooth and stable, even if the first vehicle experiences a sudden fault or unexpected situation, the intelligent driving system or driver has more time to detect and resolve it. However, if the first vehicle is in a high-uncertainty state such as high speed, high acceleration, rapid steering, or has a fault, an erroneous, inaccurate, or delayed control command at a certain time step will greatly increase the probability of a risk. Compared to risk data, vehicle driving data is constantly detected by the vehicle itself during driving and usually does not require additional calculations, thus eliminating the need for additional detection and computing resources. Therefore, monitoring vehicle driving information first, and then performing risk data detection only after determining that the first vehicle has a potential risk based on the vehicle driving information, can effectively save resources. The specific methods and conditions for determining whether the first vehicle has a potential risk based on vehicle driving information can be determined according to actual conditions and actual vehicle calibration results, etc., and this embodiment does not impose any restrictions on this.

[0037] As an example, step S10 includes: after the first vehicle is powered on and running, continuously monitoring whether the first vehicle enters the intelligent driving state, and after detecting that the first vehicle has entered the intelligent driving state, obtaining the current vehicle driving information of the first vehicle.

[0038] Step S20: If the vehicle driving information is determined to meet the preset potential risk conditions, risk data detection is performed.

[0039] As an example, step S20 includes: determining whether the vehicle driving information meets preset potential risk conditions; if the vehicle driving information meets the preset potential risk conditions, then it is determined that the first vehicle has a potential risk, and then one or more risk items are further detected to obtain the risk data corresponding to each risk item. For example, if the risk item is acceleration, then the risk data is the specific value of acceleration; if the vehicle driving information does not meet the preset potential risk conditions, then risk data detection is not required, and the process can return to the step of obtaining vehicle driving information when it is determined that the first vehicle is in intelligent driving state to continue detecting whether the first vehicle has a potential risk. The preset potential risk conditions can be determined based on actual vehicle calibration results, etc., and this embodiment does not impose any restrictions on this.

[0040] In one embodiment, the risk data can be the sensing data itself or the difference between sensing data generated at different times or by different sensors.

[0041] In another embodiment, risk data can be assigned based on perception data. For example, if the distance between the target object detected by the camera and the first vehicle is less than or equal to a preset threshold, the risk data is set to 1; if the distance between the target object detected by the camera and the first vehicle is greater than the preset threshold, the risk data is set to 0. If the category of the target object detected by the radar is different from the category of the target object detected by the camera, the risk data is set to 1; if the category of the target object detected by the radar is the same as the category of the target object detected by the camera, the risk data is set to 0. In this case, the corresponding high-risk warning condition can be set to greater than 0. Then, if the distance between the target object and the first vehicle is less than or equal to the preset threshold, or if the category of the target object detected by the radar is different from the category of the target object detected by the camera, the corresponding high-risk warning condition is met, and intelligent driving risk warning information will be output. In this way, if it is necessary to calculate or statistically analyze different types of risk data, the computational load is lower and the efficiency is higher.

[0042] The vehicle driving information includes at least one of vehicle acceleration, vehicle deceleration, and steering speed. The step of performing risk data detection when the vehicle driving information meets preset potential risk conditions includes:

[0043] If it is detected that there is target vehicle driving information that meets the preset potential risk conditions in the vehicle driving information, the risk data corresponding to the target vehicle driving information is detected, wherein the preset potential risk conditions include at least one of the following: the vehicle acceleration exceeds the preset second acceleration threshold, the vehicle deceleration exceeds the preset second deceleration threshold, or the steering speed exceeds the preset second steering speed threshold.

[0044] In this embodiment, since the driving state of the first vehicle varies, the risk points during the driving process are different, and therefore the risk data to be detected also differs. For example, during acceleration, risk points may focus on the distance between the vehicle and the vehicle in front, and whether there is an intersection ahead, while during deceleration, risk points may focus on the speed of the vehicles behind, and whether there are any road users suddenly cutting in or out. Therefore, a mapping relationship between driving state and risk items can be established in advance. When it is determined that the target vehicle's driving information meets the preset potential risk conditions, the current target driving state of the first vehicle can be determined based on the target vehicle's driving information. According to the preset mapping relationship between driving state and risk items, the target risk item corresponding to the target driving state can be determined, and the target risk item can be detected to obtain the risk data required to satisfy the current driving state of the first vehicle.

[0045] Potential risk conditions include at least one of the following: the vehicle's acceleration exceeding a preset second acceleration threshold, the vehicle's deceleration exceeding a preset second deceleration threshold, or the steering speed exceeding a preset second steering speed threshold. The second acceleration threshold, second deceleration threshold, and second steering speed threshold can all be determined based on actual conditions and vehicle calibration results, and this embodiment does not impose any restrictions on them.

[0046] As an example, it is determined whether the driving information of each vehicle meets the preset potential risk conditions. If the driving information of each vehicle meets the preset potential risk conditions, the current target driving state of the first vehicle is determined according to the target vehicle driving information. According to the preset mapping relationship between driving state and risk item, the target risk item corresponding to the target driving state is determined. The target risk item is detected to obtain the risk data required to meet the current driving state of the first vehicle.

[0047] Step S30: If a risk assessment is performed based on the risk data using a preset risk assessment model, and it is determined that the first vehicle has a risk of intelligent driving, then intelligent driving risk warning information is output.

[0048] In this embodiment, since the detection process of each risk data and the judgment process of whether the corresponding high-risk warning condition is met can be performed independently, they can be executed in parallel or in a queue. Therefore, the time to obtain the judgment result may be different. After detecting a preset number of target risk data that meet the preset high-risk warning condition, it is not necessary to continue consuming computing power to detect and judge other risk data, thereby saving computing resources. Furthermore, it is not necessary to wait until all risk data has been detected before issuing a risk warning, thus improving the efficiency of risk warning. In one embodiment, the preset number of data can be 1, that is, a high-risk warning is issued when any risk data that meets the corresponding high-risk warning condition is detected, thereby ensuring the safety of intelligent driving.

[0049] The risk assessment model is used to evaluate the riskiness of its corresponding risk data. It can be a logical model or a neural network model, etc., which can be determined according to the actual situation. This embodiment does not impose any restrictions on this. For example, if the risk item is acceleration, the risk assessment model can determine whether the acceleration is higher than a preset acceleration threshold. If the acceleration is higher than the preset acceleration threshold, it is determined that the first vehicle has a risk of intelligent driving. If the acceleration is not higher than the preset acceleration threshold, it is determined that the first vehicle does not have a risk of intelligent driving. As another example, if the risk item is environmental perception information, the risk assessment model can be a binary classification model. The binary classification model extracts feature information from the environmental perception information and calculates the confidence level of the first vehicle having a risk of intelligent driving based on the feature information. If the confidence level is higher than a preset confidence threshold, the classification result of the first vehicle having a risk of intelligent driving is output. If the confidence level is not higher than the preset confidence threshold, the classification result of the first vehicle not having a risk of intelligent driving is output.

[0050] Intelligent driving risk warning information can be in the form of images, text, audio, etc. For example, intelligent driving risk warning images or text can be output through the display on the first vehicle, or intelligent driving risk warning audio can be output through the speaker on the first vehicle. The specific form can be determined according to the actual situation, and this embodiment does not limit it.

[0051] In one embodiment, an HMI (Human Machine Interface) alert area can be set on the display device on the first vehicle. This HMI alert area outputs intelligent driving risk warning information and provides guidance on risk mitigation. The HMI alert area can be designed as a robot-like interactive text display or equipped with dynamic interactive alert icons. Through more reasonable human-machine interaction, the driver can fully understand the actual situation of the intelligent driving system and have sufficient time to check and mitigate risks, thus making intelligent driving safer and more reliable, providing a more confident and better driving experience. Simultaneously, by designing an intelligent driving interaction experience similar to robot communication, the vehicle becomes more agile, like a living travel companion, with a more complete human-machine integration, making it suitable for advanced intelligent driving.

[0052] As an example, step S30 includes: inputting each of the risk data into a preset risk assessment model; determining whether the first vehicle has an intelligent driving risk based on the risk data using the preset risk assessment model; and outputting intelligent driving risk warning information if the preset risk assessment model determines that the first vehicle has an intelligent driving risk. If the preset risk assessment model determines that the first vehicle does not have an intelligent driving risk, it is not necessary to output intelligent driving risk warning information. Instead, the process can return to the step of obtaining vehicle driving information when the first vehicle is determined to be in an intelligent driving state, continuously monitoring the first vehicle to ensure the timeliness of the intelligent driving risk warning.

[0053] In one embodiment, if a preset number of target risk data that meet preset high-risk warning conditions are detected from the various risk data, a prompt message to deactivate intelligent driving operation can also be output. For example, deactivating intelligent driving during steering can be done by gripping the steering wheel tightly, and the operation can be displayed on the first vehicle with an image or video of gripping the steering wheel tightly, or the operation can be output through the first vehicle's speaker with an audio message of gripping the steering wheel tightly; deactivating intelligent driving during acceleration can be done by pressing the brake pedal, and the operation can be displayed on the first vehicle with an image or video of pressing the brake pedal, or the operation can be output through the first vehicle's speaker with an audio message of pressing the brake pedal; deactivating intelligent driving during deceleration can be done by pressing the accelerator pedal or pressing the button to deactivate intelligent driving control, and the operation can be displayed on the first vehicle with an image or video of pressing the accelerator pedal and pressing the button to deactivate intelligent driving control, or the operation can be output through the first vehicle's speaker with an audio message of pressing the accelerator pedal and pressing the button to deactivate intelligent driving control.

[0054] In one embodiment, due to the different driving states of the first vehicle, the risk points during the driving process are different, and therefore the risk data to be detected are also different, as are the risk assessment models used. For example, during acceleration, risk points may be concentrated on the distance between the vehicle and the vehicle in front, and whether there is an intersection ahead, while during deceleration, risk points may be concentrated on the speed of the vehicle behind, and whether there are traffic participants suddenly cutting in or out ahead. Therefore, different assessment methods can be selected for the first vehicle under different driving states. Thus, a mapping relationship between driving state, risk items, and risk assessment model can be established in advance. When it is determined that the vehicle driving information meets the preset potential risk conditions, the current target driving state of the first vehicle can be determined based on the vehicle driving information. According to the preset mapping relationship between driving state, risk items, and risk assessment model, the target risk item and target risk assessment model corresponding to the target driving state are determined, the target risk item is detected, the target risk data is obtained, and then the risk assessment is performed based on the target risk data using the target risk assessment model.

[0055] The risk assessment model includes an overtaking and lane-changing risk assessment model, and the risk data further includes at least one of overtaking and lane-changing instructions and steering speed; the step of conducting risk assessment based on the risk data using a preset risk assessment model includes:

[0056] By using a preset overtaking and lane-changing risk assessment model, it is determined whether there are more than a preset fourth number of risk data that meet the preset overtaking and lane-changing risk warning conditions. The preset overtaking and lane-changing risk warning conditions include at least one of: detecting an overtaking and lane-changing command and the steering speed exceeding a preset first steering speed threshold.

[0057] In this embodiment, the driving state of the first vehicle includes an overtaking and lane-changing state. The risk assessment model corresponding to the overtaking and lane-changing state is an overtaking and lane-changing risk assessment model. The risk data corresponding to the overtaking and lane-changing state includes at least one of an overtaking and lane-changing command and a steering speed. The safety risks during overtaking and lane-changing are generally high. Therefore, for the overtaking and lane-changing state, it is only necessary to verify the accuracy of the overtaking and lane-changing state. Once it is determined that the vehicle is in the overtaking and lane-changing state, intelligent driving risk warnings can be continuously issued. The overtaking and lane-changing command refers to the command issued by the intelligent driving system or the controller to control the vehicle to overtake and change lanes. Whether the first vehicle is currently in the overtaking and lane-changing state can be verified by detecting whether the overtaking and lane-changing command and / or the steering speed exceeds a preset first steering speed threshold.

[0058] As an example, overtaking and lane-changing commands and / or steering speeds are input into a preset overtaking and lane-changing risk assessment model. The model determines whether an overtaking and lane-changing command exists in the input and whether the input steering speed exceeds a preset first steering speed threshold. If an overtaking and lane-changing command is detected in the input, it is determined that the command meets the overtaking and lane-changing risk warning conditions; otherwise, it does not. If the steering speed exceeds the preset first steering speed threshold, it is determined that the steering speed meets the preset overtaking and lane-changing risk warning conditions; otherwise, it does not. The number of risk data points meeting the preset overtaking and lane-changing risk warning conditions is determined, and it is determined whether the number of such risk data points is greater than or equal to a preset fourth number.

[0059] The risk assessment model includes an acceleration risk assessment model, and the risk data further includes at least one of vehicle acceleration and first external environmental information; the step of conducting risk assessment based on the risk data using a preset risk assessment model includes:

[0060] By accelerating the risk assessment model, it is determined whether there are more than a preset first number of risk data that meet the preset acceleration risk warning conditions. The preset acceleration risk warning conditions include: the perception confidence exceeds a preset first perception confidence threshold, the fusion confidence exceeds a preset first fusion confidence threshold, the vehicle acceleration exceeds a preset first acceleration threshold, and at least one of the following is identified based on the first external environment information: at least one intersection is identified within a preset first distance range in front of the first vehicle.

[0061] In this embodiment, the risk data includes acceleration risk data, which refers to data that may cause risks to the first vehicle during acceleration. This data may include one or more of the following: vehicle status information, external environment information, and intelligent driving verification information. Vehicle status information refers to the status information of the first vehicle, including vehicle acceleration and steering wheel angle. For example, if there are abnormalities in the acceleration sensor or vehicle speed sensor, or if there is a delay in data transmission, and the intelligent driving decision results in excessive acceleration or high acceleration during a sharp turn, it can easily lead to a safety accident. External environment information refers to information about the environment in which the first vehicle is located, including road condition information, lane reference lines, and obstacle information. Obstacles include pedestrians, traffic facilities, and other vehicles besides the first vehicle. Obstacle information includes obstacle location and obstacle speed. When the intelligent driving decision violates the environmental safety intent, such as when there is an intersection or sharp turn nearby, continued acceleration can easily lead to a safety accident. Intelligent driving verification information refers to relevant information used to verify the accuracy of intelligent driving decisions. This includes the perception confidence level of each perception module's independent perception of the external environment, the fusion confidence level characterizing the deviations between the perception capabilities of various modules, and sensor accuracy. For example, a vehicle equipped with radar and cameras may encounter pedestrians and trucks ahead. If the truck passes in front of the pedestrian, it may obscure the pedestrian's view, causing the pedestrian to be undetectable at a certain moment. If the intelligent driving system accelerates the vehicle based on the perception data of this undetected pedestrian, it could lead to a safety accident. Similarly, if there are delays in camera data transmission or image recognition, the pedestrian's position perceived by the camera may differ significantly from that perceived by the radar. In this case, if the intelligent driving system accelerates the vehicle based on the pedestrian's position perceived by the camera, it could also lead to a safety accident. The perception accuracy of the perception modules is a crucial factor affecting intelligent driving decisions. However, environmental perception technology, sensor accuracy, and data transmission delays can all affect the accuracy of the perception modules' perception of the external environment, thus impacting the accuracy of intelligent driving decisions.

[0062] Different risk data correspond to different acceleration risk warning conditions. The acceleration risk warning condition corresponding to perception confidence is that the perception confidence exceeds a preset first perception confidence threshold. The acceleration risk warning condition corresponding to fusion confidence is that the fusion confidence exceeds a preset first fusion confidence threshold. The acceleration risk warning condition corresponding to vehicle acceleration is that the vehicle acceleration exceeds a preset first acceleration threshold. The acceleration risk warning condition corresponding to first external environment information is that, based on the first external environment information, at least one of at least one intersection is identified within a preset first distance range in front of the first vehicle.

[0063] As an example, by using an accelerated risk assessment model, it is determined whether each of the aforementioned risk data meets its corresponding accelerated risk warning conditions, and then it is determined whether the number of risk data that meets the preset accelerated risk warning conditions is greater than a preset first number.

[0064] The risk assessment model includes a deceleration risk assessment model, and the risk data also includes the vehicle's deceleration; the step of conducting a risk assessment based on the risk data using a preset risk assessment model includes:

[0065] The deceleration risk assessment model determines whether there are more than a preset second number of risk data that meet the preset deceleration risk warning conditions. The preset deceleration risk warning conditions include: the perception confidence exceeds a preset second perception confidence threshold, the fusion confidence exceeds a preset second fusion confidence threshold, and the vehicle deceleration exceeds a preset first deceleration threshold.

[0066] In this embodiment, deceleration risk data refers to data that may cause risks to the first vehicle during deceleration, and may include one or more of the following: vehicle status information, external environment information, and intelligent driving verification information. Vehicle status information refers to the status information of the first vehicle, including vehicle deceleration, steering wheel angle, etc. For example, if there are abnormalities in the deceleration sensor or vehicle speed sensor, or if there is a delay in data transmission, resulting in excessive vehicle deceleration or high vehicle deceleration during a sharp turn, intelligent driving decisions can easily lead to safety accidents. Furthermore, scenarios with high vehicle deceleration are inherently riskier. For instance, if a vehicle suddenly cuts in front, a greater vehicle deceleration is required to avoid a collision. Or, while following another vehicle, if the vehicle in front suddenly cuts out to avoid a pedestrian, and the pedestrian is obscured by the vehicle in front and not detected by the vehicle, the vehicle in front may be at risk after the vehicle in front cuts out. Vehicles require significant deceleration to avoid hitting pedestrians. In these scenarios with high deceleration, if a system malfunction causes the first vehicle to exit autonomous driving mode and the deceleration control command to cease, the first vehicle will no longer decelerate. If the driver's attention is not focused on the vehicle's movement, they may not notice the abnormality. By the time the driver notices it, it may be too late to apply the brakes, resulting in a high risk of collision. If a deceleration risk warning is issued when entering these scenarios, the driver can prepare to take over the vehicle in advance and focus on assessing whether the vehicle's driving state is reasonable. In this way, if a system malfunction causes the vehicle to exhibit abnormal driving behavior that violates the driver's intentions or environmental safety requirements, the driver can take over the vehicle more quickly and promptly, thereby reducing safety risks. External environmental information refers to information about the environment in which the first vehicle is located, including road condition information, lane reference lines, obstacle information, etc. Among them, obstacles include pedestrians, traffic facilities, and other vehicles besides the first vehicle. Obstacle information includes obstacle location, obstacle speed, etc. When the intelligent driving decision violates the environmental safety intention, such as when driving on a highway, when the speed of the vehicle behind is high, or when the distance between the vehicle behind and the vehicle is small, continuing to decelerate can easily lead to a safety accident.Intelligent driving verification information refers to relevant information used to verify the accuracy of intelligent driving decisions. This includes the perception confidence of each perception module's independent perception of the external environment, the fusion confidence representing the deviation between the perception capabilities of various perception modules, and sensor accuracy. For example, a vehicle equipped with radar and cameras may encounter a slow-moving truck behind it on the right rear, with a car behind the truck traveling at high speed and intending to change lanes. The truck may obstruct the car at times, potentially preventing the system from detecting or recognizing the car's lane-changing intention. In this situation, the intelligent driving system's deceleration decision could lead to a safety accident. Similarly, delays in camera data transmission or image recognition can cause significant differences between the camera's perception of obstacle vehicles and the radar's perception of pedestrians. If the intelligent driving system decelerates based solely on the camera's perception of the pedestrian's position, it could also lead to a safety accident. The perception accuracy of the perception modules is a crucial factor affecting intelligent driving decisions. However, environmental perception technology, sensor accuracy, and data transmission delays can all affect the accuracy of the perception modules' perception of the external environment, thus impacting the accuracy of intelligent driving decisions.

[0067] Different risk data correspond to different deceleration risk warning conditions. The deceleration risk warning condition corresponding to the perception confidence level is that the perception confidence level exceeds the preset second perception confidence level threshold. The deceleration risk warning condition corresponding to the fusion confidence level is that the fusion confidence level exceeds the preset second fusion confidence level threshold. The deceleration risk warning condition corresponding to the vehicle deceleration is that the vehicle deceleration exceeds the preset first deceleration level threshold.

[0068] As an example, the deceleration risk assessment model is used to determine whether each of the aforementioned risk data meets its corresponding deceleration risk warning conditions, and then to determine whether the number of risk data that meets the preset deceleration risk warning conditions is greater than a preset second number.

[0069] The risk assessment model includes a steering risk assessment model, and the risk data further includes at least one of road information, environmental perception data, and actual steering direction; the step of conducting risk assessment based on the risk data using a preset risk assessment model includes:

[0070] Obtain the specified steering direction for intelligent driving;

[0071] The steering risk assessment model determines whether there are more than a preset third number of risk data that meet preset steering risk warning conditions. The preset steering risk warning conditions include at least one of the following: the perception confidence exceeds a preset third perception confidence threshold; the fusion confidence exceeds a preset third fusion confidence threshold; at least one target is identified within a preset second distance range on the left and right sides of the first vehicle based on the environmental perception data; at least one steering risk road segment is identified within a preset third distance range from the first vehicle based on the road information; and the actual steering direction is different from the steering direction specified by the intelligent driving system. The steering risk road segment includes at least one of the following: intersections, complex lane sections, and sharp turns.

[0072] In this embodiment, steering risk data refers to data that may cause risks to the first vehicle during steering, and may include one or more of the following: vehicle status information, environmental perception information, and intelligent driving verification information. Vehicle status information refers to the status information of the first vehicle, including steering speed and vehicle speed. For example, if there is an anomaly in the steering speed sensor or vehicle speed sensor, or if there is a delay in data transmission, intelligent driving decisions such as excessive steering speed or excessive steering angle can easily lead to safety accidents. Environmental perception information refers to information about the environment in which the first vehicle is located, including road information, lane reference lines, and obstacle information. Obstacles include pedestrians, traffic facilities, and other vehicles besides the first vehicle. Obstacle information includes obstacle position and obstacle speed. When intelligent driving decisions violate environmental safety intentions, such as when there is an intersection nearby or the vehicle is making a sharp turn, inaccurate steering timing, excessively large or small steering angles can easily lead to collisions. Intelligent driving verification information refers to relevant information used to verify the accuracy of intelligent driving decisions. This includes the perception confidence of each perception module's independent perception of the external environment, the fusion confidence representing the deviation between the perception capabilities of various perception modules, and sensor accuracy. For example, a vehicle equipped with radar and cameras may need to turn right at an intersection, but there may be pedestrians crossing the road. Sometimes, these pedestrians may be obscured by other vehicles or pedestrians, making them undetectable. If the intelligent driving system controls the vehicle's steering based on perception data that fails to detect the pedestrian, it could lead to a safety accident. Similarly, delays in camera data transmission or image recognition can cause significant differences between the pedestrian's position perceived by the camera and that perceived by the radar. In such cases, if the intelligent driving system controls the vehicle's steering based on the pedestrian's position perceived by the camera, it could also lead to a safety accident. The perception accuracy of the perception modules is a crucial factor affecting intelligent driving decisions. However, environmental perception technology, sensor accuracy, and data transmission delays can all affect the accuracy of the perception modules' perception of the external environment, thus impacting the accuracy of intelligent driving decisions.

[0073] Road information refers to road information within a certain range of the first vehicle, which may include road type, road width, lane width, etc. Obstacles include pedestrians, traffic facilities, and other vehicles besides the first vehicle. Obstacle information includes obstacle location, obstacle speed, etc. The road information can be detected by determining it based on high-precision maps and positioning information, or by sensing it through a perception module on the first vehicle.

[0074] Environmental perception data refers to information about the environment in which the first vehicle is located, perceived by the perception module on the first vehicle. This includes road condition information, lane reference lines, obstacle information, etc. Obstacles include pedestrians, traffic facilities, and other vehicles besides the first vehicle. Obstacle information includes obstacle location and speed. Environmental perception data can be provided by the perception module. For example, it can be an image of the external environment captured by a camera, or it can be an external environment model generated by integrating information from multiple perception modules such as cameras and radar. The specific method can be determined according to the actual situation, and this embodiment does not impose any limitations. The external environment has a significant impact on driving safety. When intelligent driving decisions violate environmental safety intentions, such as when there are other vehicles in the left and right lanes, or when vehicles are merging into an intersection ahead, inaccurate steering timing, excessively large or small steering angles can easily lead to collisions.

[0075] The actual steering direction refers to the direction in which the first vehicle is steered by its steering controller, which can be detected by a steering direction sensor. The designated steering direction for intelligent driving refers to the steering direction in the steering decision made by the intelligent driving system. After making the steering decision, the intelligent driving system sends a control command to the controller. However, the opposite direction in the steering direction control command can be represented by positive and negative signs, for example, right is positive and left is negative. Thus, before the controller issues the final control command, data transmission delays, system malfunctions, or other reasons may cause a discrepancy between the final control command and the steering decision made by the intelligent driving system. For example, the road ahead may actually be a right curve, but the controller issues a left turn command. If the driver has to realize the mistake and correct the vehicle, in situations where time is limited, a collision with a curb is inevitable. Currently, most vehicles are equipped with steering direction sensors that can detect the actual steering direction of the vehicle. The designated steering direction for intelligent driving is generated when the intelligent driving system makes its decision. Therefore, the actual steering direction and the designated steering direction for intelligent driving can be obtained quickly and easily. Comparing the two allows for rapid verification of the correctness of the steering direction, which is a low-cost and effective method to improve the safety of intelligent driving.

[0076] Different risk data correspond to different steering risk warning conditions. The steering risk warning condition corresponding to perception confidence is that the perception confidence exceeds a preset third perception confidence threshold. The steering risk warning condition corresponding to fusion confidence is that the fusion confidence exceeds a preset third fusion confidence threshold. The steering risk warning condition corresponding to environmental perception data is that at least one target object is identified within a preset second distance range on the left and right sides of the first vehicle based on the environmental perception data. The steering risk warning condition corresponding to road information is that at least one steering risk road segment is identified within a preset third distance range from the first vehicle based on the road information. The steering risk warning condition corresponding to steering direction is that the actual steering direction is different from the steering direction specified by the intelligent driving system.

[0077] The turning-risk road sections include at least one of the following: intersections, complex lane lines, and sharp turns. Intersections include merging and merging traffic conditions. Complex lane lines refer to lane lines that differ from the standard lane lines. For example, a standard lane line has the correct lane width and two solid or dashed lines on each side, or lines characteristic of normal traffic regulations. However, due to detours, road construction, or other reasons, three or more lane lines may be identified, and the width between any two adjacent lane lines may not be equal to the lane width. Sharp turns refer to road sections with a curve radius smaller than a preset radius threshold. The curve radius can be determined based on high-precision maps and positioning information, or sensed by a perception module on the first vehicle. In one embodiment, a sharp turn may refer to a curve with a radius less than 250m. The road section that the vehicle is currently or about to travel on has a significant impact on driving safety. When intelligent driving decisions violate the environmental safety intent, such as when there is an intersection not far ahead, the current road has complex lane markings, or the vehicle is currently making a sharp turn, it is easy for the intelligent driving decision to be inaccurate. Inaccurate steering timing, excessive or insufficient steering angle can all easily lead to collision accidents.

[0078] As an example, by using a risk assessment model, it is determined whether each of the aforementioned risk data meets its corresponding risk warning conditions, and then it is determined whether the number of risk data that meets the preset risk warning conditions is greater than a preset second number.

[0079] In one embodiment, referring to FIG2, the intelligent driving risk warning device includes an input module deployed in the intelligent driving cruise system, a risk model deployed in the ECU (Electronic Control Unit), and a human-machine interaction display module. The input module includes a deceleration signal input module, an acceleration signal input module, a left / right turn signal input module, and an overtaking / lane change signal input module, which are respectively used to receive vehicle driving information. When the vehicle driving information input by any one or more of the input modules meets the preset potential risk conditions, risk data is detected, and the risk model performs risk estimation during deceleration, acceleration, left / right turn, and overtaking / lane change based on the risk data. Furthermore, when any one or more target risk models among the risk models estimate that the first vehicle has an intelligent driving risk, the target risk is input through the human-machine interaction display module. For example, the intelligent driving risk warning information corresponding to the model can output the following warnings: When there is an intelligent driving risk during deceleration, the warning message "During intelligent driving deceleration, you can press the accelerator or press the cancel button to release the brake" can be output; when there is an intelligent driving risk during acceleration, the warning message "During intelligent driving acceleration, you can press the brake to release the acceleration" can be output; when there is an intelligent driving risk during left / right turning, the warning message "During intelligent driving left / right turning, you can firmly grip the steering wheel to release the steering" can be output; and when there is an intelligent driving risk during overtaking / lane changing, the warning message "During intelligent driving overtaking / lane changing, you can press the brake and firmly grip the steering wheel to release control" can be output.

[0080] In this embodiment, the intelligent driving risk warning method is applied to a first vehicle with intelligent driving capabilities. First, by determining that the first vehicle is in an intelligent driving state, vehicle driving information is acquired, achieving monitoring of the first vehicle in this state. Then, by determining that the vehicle driving information meets preset potential risk conditions, risk data detection is performed, achieving risk monitoring when the first vehicle has potential risks. Finally, by performing a risk assessment based on the risk data using a preset risk assessment model, and determining that the first vehicle has intelligent driving risks, intelligent driving risk warning information is output, achieving timely risk warning. Thus, if intelligent driving decisions violate the driver's intentions or environmental safety intentions due to system limitations, the human driver can promptly identify and intervene to control the vehicle after receiving the intelligent driving risk warning information, achieving the purpose of risk avoidance. Therefore, it overcomes the technical defects of intelligent driving systems themselves, which may make decision-making errors due to accuracy or errors, resulting in incorrect actions or loss of control, and thus still have high safety risks due to system limitations. This method compensates for the safety loopholes in intelligent driving systems due to limitations, improving the safety of intelligent driving processes.

[0081] Example 2

[0082] Referring to Figure 3, based on the above embodiments of this application, in the second embodiment of this application, the same or similar content as the above embodiments can be referred to the above description, and will not be repeated hereafter. Based on this, the risk data includes at least one of perceived confidence and fused confidence; the step of performing risk data detection includes:

[0083] Step A10: Obtain first target information of the target object in multiple consecutive time steps, and determine the perception confidence based on the differences between each first target information.

[0084] In this embodiment, the first vehicle further includes an intelligent driving system. The intelligent driving system includes at least one perception module, such as radar or a camera. Each perception module independently perceives targets in the external environment and generates target information, which the intelligent driving system uses to make intelligent driving decisions. However, system errors, sensor accuracy, and data transmission delays in perception technology can all affect the accuracy of the intelligent driving system's decisions. When the vehicle's direction and speed change is small and road conditions are good, the intelligent driving system itself is capable of promptly correcting its initial erroneous decisions. However, the greater the vehicle's speed, acceleration, deceleration, or steering speed, the greater the changes the vehicle undergoes in a short period. The greater the impact of the intelligent driving decision deviation on driving safety, the higher the safety risk may be before the intelligent driving system corrects its erroneous decisions.

[0085] There may be one or more targets around the first vehicle. A target can refer to any single target or any target identified by the perception module that contributes to the intelligent driving decision of the first vehicle. For each target, step A10 can be executed separately to detect and determine the perception confidence, or step A20 can be executed separately to detect and determine the fusion confidence.

[0086] The first target information refers to the result of the perception module identifying the target, which may include at least one of the following: target detection information, category, speed, and specifications. The detection information refers to whether the target is identified. Targets include obstacles, traffic facilities, road reference lines, etc.

[0087] Target lane line information refers to the result of the perception module identifying the target lane line, which may include at least one of the following: target lane line detection information, category, specifications, etc., where the detection information refers to whether the target lane line has been identified.

[0088] The step size of the time step can be set according to actual conditions, such as 100ms, 200ms, 400ms, etc., and this embodiment does not limit it. In one embodiment, a series of consecutive time steps may include the current time step and / or a time step used to make intelligent driving control decisions. The first target information of the current time step can be used to judge the current perception stability and accuracy, and the first target information of the time step used to make intelligent driving control decisions can be used to judge the accuracy of the intelligent driving control decisions.

[0089] Perception confidence is used to characterize the changes in the perception results of the perception module regarding the target object and target lane line during a period of tracking. Examples include: the target object being initially detected, then suddenly lost, then suddenly detected again, and so on; the target lane line being initially detected, then suddenly lost, then suddenly detected again, and so on; the category of the identified target object changing abruptly from A to B in a later time step; the category of the identified target lane line changing abruptly from C to D in a later time step; the speed of the identified target object changing abruptly from stationary to above 30 km / h, or from above 30 km / h to stationary; the size of the identified target object suddenly increasing or decreasing by a factor of 1.5. In these cases, it indicates that the stability and accuracy of the perception module may be low during this period, and continuing intelligent driving may pose a high safety risk. Therefore, intelligent driving risk warnings are needed to remind the driver to pay attention to vehicle dynamics and decide whether to intervene in vehicle control.

[0090] In one embodiment, the perception confidence level can be calculated based on the first target information and / or target lane information at multiple time steps. For example, the first target information at multiple time steps can be subtracted pairwise, and the average, median, and maximum values ​​of the differences can be calculated as the perception confidence level corresponding to the target. Alternatively, the target lane information at multiple time steps can be subtracted pairwise, and the average, median, and maximum values ​​of the differences can be calculated as the perception confidence level corresponding to the target lane. In this case, the preset perception doubt threshold can be determined based on empirical values, real vehicle calibration results, etc., and this embodiment does not impose any restrictions on this.

[0091] In another embodiment, the perception confidence level can also be assigned based on the differences between the first target information at multiple time steps and / or the differences between the target lane information at multiple time steps. For example, if the first target information at multiple time steps is the same, or the difference, variance, etc. are less than a preset threshold, it indicates that the perception module has high perception stability and accuracy and no sudden change has occurred. Therefore, the perception confidence level is assigned a value of 0. If the first target information at multiple time steps is different, or the difference, variance, etc. exceed a preset threshold, it indicates that the perception module has low perception stability and accuracy and a sudden change has occurred. Therefore, the perception confidence level is assigned a value of 1. In this case, the preset perception doubt threshold can be set to 0.

[0092] As an example, step A10 includes: obtaining first target information for multiple consecutive time steps from the perception module or the corresponding memory, for example, obtaining first target information for the current time step and N1 time steps prior to the current time step, or obtaining first target information between the time step used to make intelligent driving control decisions and the current time step; and then determining the perception confidence level based on the differences between each first target information.

[0093] Referring to Figure 6, in one embodiment, the first target information includes at least one of obstacle detection information, first obstacle category, first obstacle speed, first obstacle specification, target lane line detection information, and target lane line category; the step of determining the perception confidence based on the differences between the various pieces of the first target information includes:

[0094] Step A11: If it is determined from the obstacle detection information that the perception module did not identify an obstacle in at least one time step, then the perception confidence corresponding to the obstacle detection information is determined to be a preset first value, wherein the preset first value is greater than a preset perception suspicion threshold.

[0095] As an example, step A11 includes: determining whether all the obstacle detection information is a target object identified. If at least one of the obstacle detection information is not a target object identified, it can be determined that the perception module did not identify the target object at at least one time step. That is, the perception module lost the target during this time, which is unstable and has a high security risk. Therefore, the perception confidence corresponding to the obstacle detection information can be determined to be a preset first value that is higher than the preset perception suspicion threshold. The first value, second value, third value, fourth value, fifth value and sixth value can be the same or different. This embodiment does not limit this.

[0096] Step A12, and / or, if at least two different obstacle categories are detected in each of the first obstacle categories, the perception confidence corresponding to the first obstacle category is determined to be a preset second value, wherein the preset second value is greater than a preset perception doubt threshold;

[0097] As an example, step A12 includes: comparing and determining whether each of the first obstacle categories is the same. If there are at least two different obstacle categories in each of the first obstacle categories, it indicates that the stability of the perception module is low during this period and there is a high security risk. Therefore, the perception confidence level corresponding to the first obstacle category can be determined as a preset second value that is higher than the preset perception suspicion threshold.

[0098] Step A13, and / or, if the difference between the speeds of the first obstacles in any two adjacent time steps exceeds a preset first speed threshold, then the perception confidence level corresponding to the speed of the first obstacle is determined to be a preset third value, wherein the preset third value is greater than a preset perception suspicion threshold.

[0099] As an example, step A13 includes: sequentially comparing the speeds of the first obstacle in two adjacent time steps, calculating the difference, and determining whether the difference exceeds a preset first speed threshold. If the difference between the speeds of the first obstacle in any two adjacent time steps exceeds the preset first speed threshold, it indicates that the stability of the perception module is low during this period and there is a high security risk. Therefore, the perception confidence level corresponding to the speed of the first obstacle can be determined as a preset third value that is higher than the preset perception suspicion threshold.

[0100] Step A14, and / or, if the difference between each of the first obstacle specifications is detected to exceed a preset first specification threshold, then the perception confidence level corresponding to the first obstacle specification is determined to be a preset fourth value, wherein the preset fourth value is greater than a preset perception suspicion threshold.

[0101] As an example, step A14 includes: calculating the difference between each of the first obstacle specifications pairwise, or comparing the speeds of the first obstacles in two adjacent time steps sequentially, and then comparing the difference with a preset first specification threshold to determine whether the difference exceeds a preset first speed threshold. If the difference between each of the first obstacle specifications is detected to exceed the preset first specification threshold, it indicates that the stability of the perception module is low during this period and there is a high safety risk. Therefore, the perception confidence corresponding to the first obstacle specification can be determined as a preset fourth value that is higher than the preset perception suspicion threshold.

[0102] Step A15, and / or, if it is determined from the target lane detection information that the perception module has not identified the target lane line at at least in at least one time step, then the perception confidence corresponding to the target lane line detection information is determined to be a preset fifth value, wherein the preset fifth value is greater than a preset perception doubt threshold.

[0103] As an example, step A15 includes: determining whether all the target lane detection information is a recognized target lane. If at least one of the target lane detection information is not a recognized target lane, it can be determined that the perception module has not recognized the target lane at at least one time step. That is, the perception module has lost the target during this time, which is unstable and poses a high security risk. Therefore, the perception confidence corresponding to the target lane detection information can be determined to be a preset fifth value that is higher than the preset perception suspicion threshold.

[0104] Step A16, and / or, if at least two different lane line categories are detected in each of the first target lane line categories, the perception confidence level corresponding to the first target lane line category is determined to be a preset sixth value, wherein the preset sixth value is greater than a preset perception doubt threshold.

[0105] As an example, step A16 includes: comparing and determining whether each of the first target lane line categories is the same. If there are at least two different lane line categories among each of the first target lane line categories, it indicates that the stability of the perception module is low during this period and there is a high safety risk. Therefore, the perception confidence level corresponding to the first target lane line category can be determined as a preset second value that is higher than the preset perception suspicion threshold.

[0106] In this embodiment, the information of the first target object perceived by the perception module at multiple time steps is compared and verified. If the differences between the information of the first target object perceived at multiple time steps are small, it indicates that the perception module is relatively stable and accurate during this period. If the differences between the information of the first target object perceived at multiple time steps are large, it indicates that the confidence level of the perception module is low during this period. Therefore, it is necessary to issue an intelligent driving risk warning to remind the driver to pay attention to the dynamics of the first vehicle, thereby avoiding incorrect intelligent driving decisions caused by sudden system failures and improving the safety of intelligent driving.

[0107] Step A20, and / or, acquiring multiple second target object information corresponding to the target object, and determining the fusion confidence level based on the differences between each second target object information, wherein each second target object information is obtained by different sensing modules sensing the target object at the same time.

[0108] In this embodiment, the second target information refers to the result of different perception modules independently identifying the target at a specified time. It may include at least one of the target's category, speed, location, and specifications. The specified time may be the current time or the time used to make intelligent driving control decisions.

[0109] When an intelligent driving system includes multiple perception modules, each perception module perceives independently. The intelligent driving system needs to fuse the perception results of each perception module to obtain a fusion result. If the perception result of one of the perception modules is inaccurate due to system limitations, it will be significantly different from the perception results of other perception modules. In this case, the fusion result is inaccurate due to the influence of the inaccurate perception result, which in turn leads to inaccurate intelligent driving decisions. Fusion confidence is used to characterize the differences between the perception results of multiple sensing modules on a target object at the same time. For example, the sensing modules may include radar and a camera. The radar identifies the target object as a pedestrian, while the camera identifies it as a vehicle; both radar and camera identify the target object as a vehicle, but the location of the target object identified by radar differs from that identified by camera by more than 30 cm; radar and camera identify the target object at the same location, but both are classified as unknown; both radar and camera identify the target object as a vehicle at the same location, but the speed of the target object identified by radar differs from that identified by camera by more than 10 km / h; both radar and camera identify the target object as a vehicle at the same location and speed, but the size of the target object identified by radar is 1.5 times that of the target object identified by camera, etc. In these cases, it indicates that the stability and accuracy of at least one sensing module may be low during this period, and continuing intelligent driving may pose a high safety risk. Therefore, intelligent driving risk warnings are needed to remind the driver to pay attention to vehicle dynamics and decide whether to intervene in vehicle control.

[0110] In one embodiment, the fusion confidence score can be calculated based on the information of each of the second target objects. For example, the information of each of the second target objects can be subtracted pairwise, and the average, median, and maximum values ​​of the differences can be calculated as the fusion confidence score. In this case, the preset fusion doubt threshold can be determined based on empirical values, actual vehicle calibration results, etc. This embodiment does not impose any restrictions on this.

[0111] In another embodiment, the fusion confidence score can also be assigned based on the differences between the information of each second target object. For example, if the information of each second target object is the same, or the difference, variance, etc. are less than a preset threshold, it indicates that the confidence score of each sensing module is high, so the fusion confidence score is assigned a value of 0. If the information of each second target object is different, or the difference, variance, etc. exceed a preset threshold, it indicates that the confidence score of at least one sensing module is low, so the fusion confidence score is assigned a value of 1. In this case, the preset perception doubt threshold can be set to 0.

[0112] As an example, step A20 includes: obtaining second target information of the target object at a specified time from the perception module or the corresponding memory, for example, obtaining the current second target information, or obtaining the second target information of the time used to make intelligent driving control decisions, etc.; and then determining the fusion confidence level based on the differences between each second target information.

[0113] Referring to Figure 7, in one embodiment, the second target information includes at least one of a second obstacle category, obstacle location, second obstacle speed, and second obstacle specification; the step of determining the fusion confidence level based on the differences between the various pieces of second target information includes:

[0114] Step A21: If at least two different obstacle categories are detected in each of the second obstacle categories, the fusion confidence level corresponding to the second obstacle category is determined to be a preset fifth value, wherein the preset fifth value is greater than a preset fusion suspicion threshold.

[0115] As an example, step A21 includes: comparing and determining whether each of the second obstacle categories is the same. If there are at least two different obstacle categories in each of the second obstacle categories, it indicates that the confidence of at least one perception module is low and there is a high security risk. Therefore, the fusion confidence corresponding to the second obstacle category can be determined as a preset fifth value that is higher than the preset fusion suspicion threshold. The fifth, sixth, seventh and eighth values ​​can be the same or different, and this embodiment does not limit this.

[0116] Step A22: If all the second obstacle categories are detected to be the same, determine whether the second obstacle category is an unknown category, and detect whether the distance between the positions of each obstacle exceeds a preset distance threshold.

[0117] Step A23: If the distance between each of the obstacle locations is detected to exceed a preset distance threshold, or if the distance between each of the obstacle locations is detected to be less than the preset distance threshold and the second obstacle category is an unknown category, then the fusion confidence level corresponding to the obstacle location is determined to be a preset sixth value, wherein the preset sixth value is greater than a preset fusion suspicion threshold.

[0118] As an example, steps A22-A23 include: after comparing and determining whether each of the second obstacle categories is the same, if it is detected that each of the second obstacle categories is the same, then it is determined whether the second obstacle category is an unknown category, and each obstacle position is compared pairwise to calculate the distance between each pair of obstacle positions. The distance between the two obstacle positions is compared with a preset distance threshold to determine whether the distance between the two obstacle positions exceeds the preset distance threshold. If it is detected that the distance between each obstacle position exceeds the preset distance threshold, or if it is detected that the distance between each obstacle position does not exceed the preset distance threshold and the second obstacle category is an unknown category, then it indicates that the confidence of at least one perception module is low and there is a high security risk. Therefore, the fusion confidence corresponding to the obstacle position can be determined as a preset sixth value that is higher than the preset fusion suspicion threshold.

[0119] Step A24: If the distance between the positions of each obstacle is not exceeded by a preset distance threshold and the category of the second obstacle is not an unknown category, then detect whether the difference between the speeds of each of the second obstacles exceeds a preset second speed threshold.

[0120] Step A25: If the difference between the speeds of each of the second obstacles is detected to exceed a preset second speed threshold, then the fusion confidence level corresponding to the speed of the second obstacle is determined to be a preset seventh value, wherein the preset seventh value is greater than a preset fusion suspicion threshold.

[0121] As an example, steps A24-A25 include: if the distance between the locations of each obstacle is not exceeded by a preset distance threshold and the second obstacle category is not an unknown category, then the speeds of each second obstacle are subtracted pairwise to determine whether the difference exceeds a preset second speed threshold. If the difference between any two second obstacle speeds exceeds the preset second speed threshold, it indicates that the stability of at least one sensing module is low and there is a high security risk. Therefore, the fusion confidence corresponding to the second obstacle speed can be determined as a preset seventh value that is higher than the preset fusion suspicion threshold.

[0122] Step A26: If the difference between the speeds of each of the second obstacles is not detected to exceed a preset second speed threshold, then detect whether the difference between the specifications of each of the second obstacles exceeds a preset second specification threshold.

[0123] Step A27: If the difference between each of the second obstacle specifications is detected to exceed the preset second specification threshold, then the fusion confidence level corresponding to the second obstacle specification is determined to be a preset eighth value, wherein the preset eighth value is greater than the preset fusion suspicion threshold.

[0124] As an example, steps A26-A27 include: if the difference between the speeds of each of the second obstacles is not detected to exceed a preset second speed threshold, then the differences between each of the second obstacle specifications are calculated pairwise to determine whether the difference exceeds a preset second specification threshold. If the difference between any two second obstacle specifications is detected to exceed the preset second specification threshold, it indicates that the stability of at least one sensing module is low and there is a high security risk. Therefore, the fusion confidence corresponding to the second obstacle specification can be determined to be a preset eighth value that is higher than the preset fusion suspicion threshold.

[0125] In this embodiment, the information of the second target object perceived by multiple sensing modules at the same time is compared and verified. If the differences between the information perceived by multiple sensing modules at the same time are small, it indicates that each sensing module is relatively stable and accurate during this period. If the differences between the information perceived by multiple sensing modules at the same time are large, it indicates that at least one sensing module has a low confidence level during this period. Perception data with low confidence level will affect the intelligent driving system's intelligent driving decisions, thus posing a high safety risk. Intelligent driving risk warning is needed to remind the driver to pay attention to the dynamics of the first vehicle, thereby avoiding erroneous intelligent driving decisions caused by sudden system failures and improving the safety of intelligent driving.

[0126] Example 3

[0127] This application embodiment also provides an intelligent driving risk warning device. Referring to FIG4, the intelligent driving risk warning device is applied to a first vehicle and includes:

[0128] The acquisition module 10 is used to acquire vehicle driving information when it is determined that the first vehicle is in an intelligent driving state;

[0129] The detection module 20 is used to perform risk data detection when it is determined that the vehicle driving information meets preset potential risk conditions;

[0130] The output module 30 is used to output intelligent driving risk warning information when the first vehicle is determined to have intelligent driving risks by performing a risk assessment based on the risk data through a preset risk assessment model.

[0131] The risk data includes at least one of perceived confidence and fused confidence; the detection module 20 is further configured to:

[0132] Acquire first target information of the target object at multiple consecutive time steps, and determine the perception confidence level based on the differences between each first target information.

[0133] And / or, acquire multiple second target object information corresponding to the target object, and determine the fusion confidence level based on the differences between each second target object information, wherein each second target object information is obtained by different sensing modules sensing the target object at the same time.

[0134] The risk assessment model includes an acceleration risk assessment model, and the risk data further includes at least one of vehicle acceleration and first external environmental information; the output module 30 is also used for:

[0135] By accelerating the risk assessment model, it is determined whether there are more than a preset first number of risk data that meet the preset acceleration risk warning conditions. The preset acceleration risk warning conditions include: the perception confidence exceeds a preset first perception confidence threshold, the fusion confidence exceeds a preset first fusion confidence threshold, the vehicle acceleration exceeds a preset first acceleration threshold, and at least one of the following is identified based on the first external environment information: at least one intersection is identified within a preset first distance range in front of the first vehicle.

[0136] The risk assessment model includes a deceleration risk assessment model, and the risk data also includes the vehicle's deceleration; the output module 30 is further used for:

[0137] The deceleration risk assessment model determines whether there are more than a preset second number of risk data that meet the preset deceleration risk warning conditions. The preset deceleration risk warning conditions include: the perception confidence exceeds a preset second perception confidence threshold, the fusion confidence exceeds a preset second fusion confidence threshold, and the vehicle deceleration exceeds a preset first deceleration threshold.

[0138] The risk assessment model includes a steering risk assessment model, and the risk data further includes at least one of road information, environmental perception data, and actual steering direction; the output module 30 is also used for:

[0139] Obtain the specified steering direction for intelligent driving;

[0140] The steering risk assessment model determines whether there are more than a preset third number of risk data that meet preset steering risk warning conditions. The preset steering risk warning conditions include at least one of the following: the perception confidence exceeds a preset third perception confidence threshold; the fusion confidence exceeds a preset third fusion confidence threshold; at least one target is identified within a preset second distance range on the left and right sides of the first vehicle based on the environmental perception data; at least one steering risk road segment is identified within a preset third distance range from the first vehicle based on the road information; and the actual steering direction is different from the steering direction specified by the intelligent driving system. The steering risk road segment includes at least one of the following: intersections, complex lane sections, and sharp turns.

[0141] The risk assessment model includes an overtaking and lane-changing risk assessment model, and the risk data further includes at least one of overtaking and lane-changing instructions and steering speed; the output module 30 is also used for:

[0142] By using a preset overtaking and lane-changing risk assessment model, it is determined whether there are more than a preset fourth number of risk data that meet the preset overtaking and lane-changing risk warning conditions. The preset overtaking and lane-changing risk warning conditions include at least one of: detecting an overtaking and lane-changing command and the steering speed exceeding a preset first steering speed threshold.

[0143] The vehicle driving information includes at least one of vehicle acceleration, vehicle deceleration, and steering speed. The detection module 20 is further used for:

[0144] If it is detected that there is target vehicle driving information that meets the preset potential risk conditions in the vehicle driving information, the risk data corresponding to the target vehicle driving information is detected, wherein the preset potential risk conditions include at least one of the following: the vehicle acceleration exceeds the preset second acceleration threshold, the vehicle deceleration exceeds the preset second deceleration threshold, or the steering speed exceeds the preset second steering speed threshold.

[0145] The intelligent driving risk warning device provided by this invention employs the intelligent driving risk warning method in the above embodiments, solving the technical problem of high safety risks caused by system limitations during intelligent driving in related technologies. Compared with the prior art, the beneficial effects of the intelligent driving risk warning device provided by this invention are the same as those of the intelligent driving risk warning method provided in the above embodiments, and other technical features in this intelligent driving risk warning device are the same as those disclosed in the methods of the above embodiments, and will not be repeated here.

[0146] Example 4

[0147] This invention provides an electronic device, which includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, which are executed by the at least one processor to enable the at least one processor to perform the intelligent driving risk warning method described in the above embodiments.

[0148] Referring now to Figure 5, a schematic diagram of the structure of an electronic device suitable for implementing embodiments of this application is shown. The electronic device in the embodiments of this application may include, but is not limited to, mobile terminals such as Bluetooth headsets, mobile phones, laptops, digital broadcast receivers, PDAs (personal digital assistants), PADs (tablet computers), PMPs (portable multimedia players), in-vehicle terminals (e.g., in-vehicle navigation terminals), and fixed terminals such as digital TVs and desktop computers. The electronic device shown in Figure 5 is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of this application.

[0149] As shown in Figure 5, an electronic device may include a processing unit (e.g., a central processing unit, a graphics processing unit, etc.) that can perform various appropriate actions and processes based on a program stored in read-only memory (ROM) or a program loaded from a storage device into random access memory (RAM). The RAM also stores various programs and arrays required for the operation of the electronic device. The processing unit, ROM, and RAM are interconnected via a bus. Input / output (I / O) interfaces are also connected to the bus.

[0150] Typically, the following systems can be connected to the I / O interface: input devices including, for example, touchscreens, touchpads, keyboards, mice, image sensors, microphones, intelligent driving speedometers, gyroscopes, etc.; output devices including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; storage devices including, for example, magnetic tapes, hard disks, etc.; and communication devices. Communication devices allow electronic devices to communicate wirelessly or wiredly with other devices to exchange arrays. Although electronic devices with various systems are shown in the figures, it should be understood that it is not required to implement or possess all the systems shown. More or fewer systems may be implemented alternatively.

[0151] According to embodiments of this application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of this application include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device, or installed from a storage device, or installed from a ROM. When the computer program is executed by a processing device, it performs the functions defined in the methods of the embodiments of this application.

[0152] The electronic device provided by this invention employs the intelligent driving risk warning method in the above embodiments, solving the technical problem of high safety risks caused by system limitations during intelligent driving in related technologies. Compared with the prior art, the beneficial effects of the electronic device provided by the embodiments of this invention are the same as those of the intelligent driving risk warning method provided in the above embodiments, and other technical features of this electronic device are the same as those disclosed in the methods of the above embodiments, and will not be repeated here.

[0153] It should be understood that various parts of this application can be implemented using hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics may be combined in any suitable manner in one or more embodiments or examples.

[0154] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

[0155] Example 5

[0156] This embodiment provides a computer-readable storage medium having computer-readable program instructions stored thereon, which are used to execute the intelligent driving risk warning method in the above embodiment.

[0157] The computer-readable storage medium provided in this embodiment of the invention may be, for example, a USB flash drive, but is not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, system, or device. The program code contained on the computer-readable storage medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (radio frequency), etc., or any suitable combination thereof.

[0158] The aforementioned computer-readable storage medium may be included in an electronic device or may exist independently without being assembled into an electronic device.

[0159] The aforementioned computer-readable storage medium carries one or more programs, which, when executed by an electronic device, cause the electronic device to: acquire vehicle driving information when it is determined that the first vehicle is in an intelligent driving state; perform risk data detection when it is determined that the vehicle driving information meets preset potential risk conditions; and output intelligent driving risk warning information when it is determined that the first vehicle has an intelligent driving risk by performing a risk assessment based on the risk data through a preset risk assessment model.

[0160] Computer program code for performing the operations of this application can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0161] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present invention. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0162] The modules described in the embodiments of this application can be implemented in software or hardware. The names of the modules do not necessarily limit the functionality of the unit itself.

[0163] The computer-readable storage medium provided by this invention stores computer-readable program instructions for executing the aforementioned intelligent driving risk warning method, thus solving the technical problem of high safety risks caused by system limitations during intelligent driving in related technologies. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided in the embodiments of this invention are the same as the beneficial effects of the intelligent driving risk warning method provided in the above embodiments, and will not be repeated here.

[0164] Example 6

[0165] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the intelligent driving risk warning method described above.

[0166] The computer program product provided in this application solves the technical problem of high safety risks caused by system limitations in intelligent driving processes in related technologies. Compared with the prior art, the beneficial effects of the computer program product provided in the embodiments of this invention are the same as the beneficial effects of the intelligent driving risk warning method provided in the above embodiments, and will not be repeated here.

[0167] The above are merely embodiments of this application and do not limit the patent scope of this application. Any equivalent structural or procedural transformations made using the content of this application's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent scope of this application.

Claims

1. A method for early warning of risks in intelligent driving, wherein, The intelligent driving risk warning method is applied to the first vehicle and includes the following steps: If it is determined that the first vehicle is in intelligent driving mode, obtain vehicle driving information; If the vehicle driving information is determined to meet the preset potential risk conditions, risk data detection is performed. If a risk assessment is conducted based on the risk data using a preset risk assessment model, and it is determined that the first vehicle has a risk of intelligent driving, an intelligent driving risk warning message is output.

2. The intelligent driving risk warning method as described in claim 1, wherein, The step of obtaining vehicle driving information when it is determined that the first vehicle is in intelligent driving mode includes: After the first vehicle is powered on and running, it is continuously monitored whether the first vehicle enters the intelligent driving state. After the first vehicle enters the intelligent driving state, the current vehicle driving information of the first vehicle is obtained.

3. The intelligent driving risk warning method as described in claim 1, wherein, The step of performing risk data detection when the vehicle driving information is determined to meet preset potential risk conditions includes: Determine whether the vehicle driving information meets the preset potential risk conditions. If the vehicle driving information meets the preset potential risk conditions, it is determined that the first vehicle has potential risks. Then, one or more risk items are further detected to obtain the risk data corresponding to each risk item.

4. The intelligent driving risk warning method as described in claim 1, wherein, The step of outputting intelligent driving risk warning information when the first vehicle is determined to have intelligent driving risks by performing a risk assessment based on the risk data using a preset risk assessment model includes: Each of the aforementioned risk data is input into a preset risk assessment model. The preset risk assessment model determines whether the first vehicle has intelligent driving risks based on each of the aforementioned risk data. If the preset risk assessment model performs a risk assessment based on the risk data and determines that the first vehicle has intelligent driving risks, intelligent driving risk warning information is output.

5. The intelligent driving risk warning method as described in claim 1, wherein, The risk data includes at least one of perceived confidence and fused confidence; The steps for conducting risk data detection include: Acquire first target information of the target object at multiple consecutive time steps, and determine the perception confidence level based on the differences between each first target information. And / or, acquire multiple second target object information corresponding to the target object, and determine the fusion confidence level based on the differences between each second target object information, wherein each second target object information is obtained by different sensing modules sensing the target object at the same time.

6. The intelligent driving risk warning method as described in claim 5, wherein, The risk assessment model includes an acceleration risk assessment model, and the risk data also includes at least one of vehicle acceleration and first external environmental information; The step of conducting risk assessment based on the risk data using a preset risk assessment model includes: By accelerating the risk assessment model, it is determined whether there are more than a preset first number of risk data that meet the preset acceleration risk warning conditions. The preset acceleration risk warning conditions include: the perception confidence exceeds a preset first perception confidence threshold, the fusion confidence exceeds a preset first fusion confidence threshold, the vehicle acceleration exceeds a preset first acceleration threshold, and at least one of the following is identified based on the first external environment information: at least one intersection is identified within a preset first distance range in front of the first vehicle.

7. The intelligent driving risk warning method as described in claim 5, wherein, The risk assessment model includes a deceleration risk assessment model, and the risk data also includes the vehicle's deceleration. The step of conducting risk assessment based on the risk data using a preset risk assessment model includes: The deceleration risk assessment model determines whether there are more than a preset second number of risk data that meet the preset deceleration risk warning conditions. The preset deceleration risk warning conditions include: the perception confidence exceeds a preset second perception confidence threshold, the fusion confidence exceeds a preset second fusion confidence threshold, and the vehicle deceleration exceeds a preset first deceleration threshold.

8. The intelligent driving risk warning method as described in claim 5, wherein, The risk assessment model includes a steering risk assessment model, and the risk data also includes at least one of road information, environmental perception data, and actual steering direction. The step of conducting risk assessment based on the risk data using a preset risk assessment model includes: Obtain the specified steering direction for intelligent driving; The steering risk assessment model determines whether there are more than a preset third number of risk data that meet preset steering risk warning conditions. The preset steering risk warning conditions include at least one of the following: the perception confidence exceeds a preset third perception confidence threshold; the fusion confidence exceeds a preset third fusion confidence threshold; at least one target is identified within a preset second distance range on the left and right sides of the first vehicle based on the environmental perception data; at least one steering risk road segment is identified within a preset third distance range from the first vehicle based on the road information; and the actual steering direction is different from the steering direction specified by the intelligent driving system. The steering risk road segment includes at least one of the following: intersections, complex lane sections, and sharp turns.

9. The intelligent driving risk warning method as described in claim 1, wherein, The risk assessment model includes an overtaking and lane-changing risk assessment model, and the risk data also includes at least one of overtaking and lane-changing instructions and steering speed; The step of conducting risk assessment based on the risk data using a preset risk assessment model includes: By using a preset overtaking and lane-changing risk assessment model, it is determined whether there are more than a preset fourth number of risk data that meet the preset overtaking and lane-changing risk warning conditions. The preset overtaking and lane-changing risk warning conditions include at least one of: detecting an overtaking and lane-changing command and the steering speed exceeding a preset first steering speed threshold.

10. The intelligent driving risk warning method as described in claim 1, wherein, The vehicle driving information includes at least one of vehicle acceleration, vehicle deceleration, and steering speed. The step of performing risk data detection when the vehicle driving information meets preset potential risk conditions includes: If it is detected that there is target vehicle driving information that meets the preset potential risk conditions in the vehicle driving information, the risk data corresponding to the target vehicle driving information is detected, wherein the preset potential risk conditions include at least one of the following: the vehicle acceleration exceeds the preset second acceleration threshold, the vehicle deceleration exceeds the preset second deceleration threshold, or the steering speed exceeds the preset second steering speed threshold.

11. The intelligent driving risk warning method as described in claim 5, wherein, The first target information includes at least one of obstacle detection information, first obstacle category, first obstacle speed, first obstacle specification, target lane line detection information, and target lane line category; The step of determining the perception confidence level based on the differences between the information of each of the first target objects includes: If it is determined, based on the obstacle detection information, that the perception module has not detected an obstacle at at least in at least one time step, then the perception confidence level corresponding to the obstacle detection information is determined to be a preset first value, wherein the preset first value is greater than a preset perception suspicion threshold. And / or, if at least two different obstacle categories are detected in each of the first obstacle categories, the perception confidence level corresponding to the first obstacle category is determined to be a preset second value, wherein the preset second value is greater than a preset perception doubt threshold; And / or, if the difference between the speeds of the first obstacles at any two adjacent time steps exceeds a preset first speed threshold, the perception confidence level corresponding to the speed of the first obstacle is determined to be a preset third value, wherein the preset third value is greater than a preset perception suspicion threshold. And / or, if the difference between each of the first obstacle specifications is detected to exceed a preset first specification threshold, then the perception confidence corresponding to the first obstacle specification is determined to be a preset fourth value, wherein the preset fourth value is greater than a preset perception suspicion threshold. And / or, if it is determined from the target lane detection information that the perception module has not identified the target lane line at at least in at least one time step, then the perception confidence level corresponding to the target lane line detection information is determined to be a preset fifth value, wherein the preset fifth value is greater than a preset perception doubt threshold. And / or, if at least two different lane line categories are detected in each of the first target lane line categories, the perception confidence level corresponding to the first target lane line category is determined to be a preset sixth value, wherein the preset sixth value is greater than a preset perception doubt threshold.

12. The intelligent driving risk warning method as described in claim 5, wherein, The second target information includes at least one of the following: second obstacle category, obstacle location, second obstacle speed, and second obstacle specification; the step of determining the fusion confidence level based on the differences between the various pieces of second target information includes: If at least two different obstacle categories are detected in each of the second obstacle categories, the fusion confidence level corresponding to the second obstacle category is determined to be a preset fifth value, wherein the preset fifth value is greater than a preset fusion suspicion threshold. If all the second obstacle categories are detected to be the same, then it is determined whether the second obstacle category is an unknown category, and whether the distance between the positions of each obstacle exceeds a preset distance threshold. If the distance between the locations of the obstacles is detected to exceed a preset distance threshold, or if the distance between the locations of the obstacles is detected to be less than the preset distance threshold and the second obstacle category is an unknown category, then the fusion confidence level corresponding to the obstacle location is determined to be a preset sixth value, wherein the preset sixth value is greater than a preset fusion suspicion threshold. If the distance between the locations of each obstacle is not exceeded by a preset distance threshold and the category of the second obstacle is not an unknown category, then it is detected whether the difference between the speeds of each of the second obstacles exceeds a preset second speed threshold. If the difference between the speeds of each of the second obstacles is detected to exceed a preset second speed threshold, then the fusion confidence level corresponding to the speed of the second obstacle is determined to be a preset seventh value, wherein the preset seventh value is greater than a preset fusion suspicion threshold. If the difference between the speeds of each of the second obstacles is not detected to exceed a preset second speed threshold, then the difference between the specifications of each of the second obstacles is detected to exceed a preset second specification threshold. If the difference between each of the second obstacle specifications is detected to exceed a preset second specification threshold, then the fusion confidence level corresponding to the second obstacle specification is determined to be a preset eighth value, wherein the preset eighth value is greater than a preset fusion suspicion threshold.

13. An intelligent driving risk warning device, wherein, The intelligent driving risk warning device is applied to the first vehicle and includes: The acquisition module is used to acquire vehicle driving information when it is determined that the first vehicle is in an intelligent driving state; The detection module is used to perform risk data detection when it is determined that the vehicle driving information meets preset potential risk conditions; The output module is used to output intelligent driving risk warning information when the first vehicle is determined to have intelligent driving risks by performing a risk assessment based on the risk data using a preset risk assessment model.

14. An electronic device, wherein, The electronic device includes: At least one processor; and, A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, which, when executed by the at least one processor, enables the at least one processor to perform the steps of the intelligent driving risk warning method according to any one of claims 1 to 12.

15. A medium, wherein, The medium is a computer-readable storage medium, on which a program for implementing the intelligent driving risk warning method is stored. The program for implementing the intelligent driving risk warning method is executed by a processor to implement the steps of the intelligent driving risk warning method as described in any one of claims 1 to 12.

16. A product, said product being a computer program product, comprising a computer program, wherein, When the computer program is executed by the processor, it implements the steps of the intelligent driving risk warning method as described in any one of claims 1 to 12.

Images