Driving risk early warning method, electronic device and vehicle

Abstract

The application provides a driving risk early warning method, an electronic device and a vehicle. By collecting cross-vehicle historical manual driving data and cross-vehicle historical intelligent driving data, cross-vehicle aggregation evaluation of the risk properties of drivers and intelligent driving systems is realized, the influence of being unable to continuously track driving habits due to being bound to a single vehicle is avoided, and the accuracy of risk evaluation is improved. Cloud data interaction is used, so that the vehicle can obtain a global risk evaluation list when driving to an intersection, the physical limitation of the field of view of the vehicle-mounted sensor is broken through, and the perception of the macro risk situation of the intersection is realized. By introducing image data and cloud list for consistency matching, the cloud data lag problem is avoided, and the authenticity and accuracy of the intersection risk situation map are ensured. Combined with the dynamic risk value and the intersection risk situation map, hierarchical early warning is performed, so that the early warning strategy is more suitable for the actual risk degree and the current vehicle control mode, and the effectiveness and user experience of early warning are significantly improved.

Description

Technical Field

[0001] This application relates to the field of intelligent driving technology, and in particular to a method for warning of driving risks, electronic devices, and vehicles. Background Technology

[0002] With the increase in car ownership, intersections have become high-incidence areas for traffic accidents. However, driver monitoring systems (DMS) and dashcams can only identify predetermined driving risks such as "running a red light," lacking the ability to predict and quantify high-risk behaviors such as "running a yellow light" and "rushing a green light," resulting in a blind spot in the quantification of driving risks. Summary of the Invention

[0003] In view of this, the purpose of this application is to propose a driving risk warning method, electronic device and vehicle, which dynamically determines the dynamic risk value based on cross-vehicle historical data, and performs graded warnings based on the dynamic risk value and the current vehicle distribution at the intersection, thereby improving the effectiveness of the warning and the user experience.

[0004] To achieve the above objectives, this application provides a method for early warning of driving risks, comprising:

[0005] Driving risk assessment is performed based on cross-vehicle historical manual driving data and cross-vehicle historical intelligent driving data to obtain the dynamic risk value of the driver's vehicle at the current moment; wherein, the cross-vehicle historical manual driving data is historical manual driving behavior data associated with the driver's identity and covering different vehicles; the cross-vehicle historical intelligent driving data is historical intelligent driving decision data associated with the vehicle identification code and intelligent driving version of each vehicle; In response to the fact that the distance between the vehicle's current position and the intersection it is about to pass through is less than or equal to a preset threshold distance, a risk assessment list corresponding to the current intersection is obtained from the cloud based on the current vehicle position and the vehicle's current driving lane; the risk assessment list is used to characterize the risk attribute set of vehicles queuing in each lane of the current intersection. Based on the risk assessment list and the image data of the current intersection collected by the acquisition device, vehicle consistency matching is performed to obtain an intersection risk situation map; the intersection risk situation map is used to characterize the dynamic risk distribution and relative spatial position relationship of each traffic participant within the current intersection range; The vehicle control mode currently used by the vehicle is classified and warned according to the dynamic risk value and the intersection risk situation map; the vehicle control mode includes manual control mode or intelligent driving mode.

[0006] Based on the same inventive concept, this application also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable by the processor, wherein the processor implements the method described above when executing the computer program.

[0007] Based on the same inventive concept, this application also provides a vehicle including the electronic equipment described above.

[0008] As can be seen from the above, the driving risk warning method, electronic device, and vehicle provided in this application can assess driving risks based on cross-vehicle historical manual driving data and cross-vehicle historical intelligent driving data, obtaining the dynamic risk value of the driver's vehicle at the current moment. When the distance between the vehicle's current position and the location of the intersection to be passed is less than or equal to a preset threshold distance, a risk assessment list corresponding to the current intersection is obtained from the cloud based on the current vehicle position and the vehicle's current driving lane. Vehicle consistency matching is performed based on the risk assessment list and image data to obtain an intersection risk situation map. A graded warning is then issued based on the dynamic risk value and the intersection risk situation map to determine the vehicle control mode currently being used. By collecting cross-vehicle historical manual driving data and cross-vehicle historical intelligent driving data, cross-vehicle aggregated assessment of the risk attributes of the driver and intelligent driving system is achieved, constructing a dynamic risk value throughout the entire lifecycle. This avoids the impact of being unable to continuously track driving habits due to being bound to a single vehicle, thus improving the accuracy of risk assessment. By leveraging cloud-based data interaction, vehicles can access a comprehensive risk assessment list as they approach intersections, overcoming the physical limitations of onboard sensor field of view and enabling the perception of macroscopic risk situations at intersections. By introducing consistent matching between image data and the cloud-based list, the system effectively eliminates data lag caused by vehicle lane changes and intersection entry / exit, ensuring the authenticity and accuracy of the intersection risk situation map. Combining the vehicle's dynamic risk value with the intersection risk situation maps of surrounding vehicles for tiered early warning systems makes the warning strategy more aligned with the actual risk level and current vehicle control mode, significantly improving the effectiveness of warnings and user experience. Attached Figure Description

[0009] To more clearly illustrate the technical solutions in this application or related technologies, the drawings used in the description of the embodiments or related technologies will be briefly introduced below. Obviously, the drawings described below are only embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0010] Figure 1 This is a flowchart of a driving risk warning method according to an embodiment of this application; Figure 2This is a schematic diagram of a driving risk warning device according to an embodiment of this application; Figure 3 This is a schematic diagram of an electronic device according to an embodiment of this application. Detailed Implementation

[0011] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with specific embodiments and the accompanying drawings.

[0012] It should be noted that, unless otherwise defined, the technical or scientific terms used in the embodiments of this application should have the ordinary meaning understood by one of ordinary skill in the art to which this application pertains. The terms "first," "second," and similar terms used in the embodiments of this application do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Terms such as "comprising" or "including" mean that the element or object preceding the word encompasses the elements or objects listed after the word and their equivalents, without excluding other elements or objects. Terms such as "connected" or "linked" are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. Terms such as "upper," "lower," "left," and "right" are only used to indicate relative positional relationships; when the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0013] It is important to understand in this article that any number of elements in the accompanying figures is for illustrative purposes and not for limitation, and that any naming is for distinction only and has no limiting meaning.

[0014] Based on the above background description, the following situations also exist in the related technologies: The risk warning and control technologies in these technologies still have significant limitations in practical applications. First, driver monitoring systems and dashcams can only record the established fact of "running a red light," lacking the ability to predict and quantify high-risk behaviors such as "running a yellow light" and "rushing a green light," making it difficult to distinguish between objective congestion and subjective malice, resulting in serious quantitative blind spots. Second, the risk assessment mechanisms in these technologies are mostly tied to vehicle identification numbers, ignoring the influence of the driver's driving style, leading to a separation between driving risk and driver identity, making continuous cross-vehicle tracking impossible, and also making it difficult to determine liability in human-machine co-driving scenarios. Third, the safety protection in these technologies is mostly single-vehicle warning, lacking collaborative perception of the risk status of surrounding vehicles, leaving intersection participants in a state of information asymmetry, which can easily trigger chain accidents. Finally, the driving score systems in these technologies are mostly static accumulations, lacking dynamic decay and incentive mechanisms, unable to be adjusted in real time based on the driver's recent performance, and failing to play an effective supervisory role. Therefore, there is an urgent need for an intersection risk assessment and warning solution that can quantify high-risk behaviors, track identities across vehicles, achieve multi-vehicle collaborative defense, and has a dynamic management mechanism.

[0015] The driving risk warning method, electronic equipment, and vehicle provided in this application can assess driving risks based on cross-vehicle historical manual driving data and cross-vehicle historical intelligent driving data, obtaining the dynamic risk value of the driver's vehicle at the current moment. When the distance between the vehicle's current position and the location of the intersection it is about to pass is less than or equal to a preset threshold distance, a risk assessment list corresponding to the current intersection is obtained from the cloud based on the vehicle's current position and current driving lane. Vehicle consistency matching is performed based on the risk assessment list and image data to obtain an intersection risk situation map. A graded warning is then issued based on the dynamic risk value and the intersection risk situation map to determine the vehicle control mode currently being used. By collecting cross-vehicle historical manual driving data and cross-vehicle historical intelligent driving data, cross-vehicle aggregated assessment of the risk attributes of the driver and intelligent driving system is achieved, constructing a dynamic risk value throughout the entire lifecycle. This avoids the impact of not being able to continuously track driving habits due to binding to a single vehicle, improving the accuracy of risk assessment. Utilizing cloud data interaction, the vehicle can obtain a full-domain risk assessment list when approaching an intersection, breaking through the physical limitations of the field of view of onboard sensors and realizing the perception of the macro-risk situation at the intersection. By introducing image data and performing consistency matching with cloud-based lists, the problem of cloud data lag caused by vehicle lane changes and intersection entry / exit is effectively eliminated, ensuring the authenticity and accuracy of the intersection risk situation map. Combining the vehicle's dynamic risk value with the intersection risk situation map of surrounding vehicles for tiered early warning makes the warning strategy more closely aligned with the actual risk level and current vehicle control mode, significantly improving the effectiveness of the warnings and the user experience.

[0016] The driving risk warning method provided by the embodiments of this application will be described in detail below with reference to the accompanying drawings.

[0017] In some embodiments, such as Figure 1 As shown, a method for warning of driving risks includes: Step 101 - Step 104.

[0018] Step 101: Based on cross-vehicle historical manual driving data and cross-vehicle historical intelligent driving data, conduct a driving risk assessment to obtain the dynamic risk value of the driver's vehicle at the current moment.

[0019] Among them, cross-vehicle historical manual driving data is historical manual driving behavior data associated with driver identity and covering different vehicles; cross-vehicle historical intelligent driving data is historical intelligent driving decision data associated with vehicle identification codes and intelligent driving versions of each vehicle.

[0020] In practical implementation, cross-vehicle historical manual driving data refers to a structured dataset reflecting a driver's manual control behavior, recorded by onboard sensors when the driver's unique FaceID, obtained through the Driver Monitoring System (DMS), is associated with all different vehicles driven by that driver within a historical statistical period (distinguished by Vehicle Identification Number, VIN). Cross-vehicle historical intelligent driving data refers to a dataset aggregating the decisions and states generated by different vehicles performing intelligent driving functions within a historical statistical period, based on the Vehicle Identification Number (VIN) and the intelligent driving version number as unique indexes.

[0021] Using cross-vehicle data is the cornerstone of ensuring accurate dynamic risk assessment. First, historical cross-vehicle data is continuously collected through an onboard sensor cluster, which includes a forward-facing camera, a DMS infrared camera, a GPS integrated navigation unit, and a Controller Area Network (CAN) bus. For manual driving data, the vehicle system uses the driver's FaceID obtained from the DMS as the driver's unique identifier. This driver's identity is then mapped to the Vehicle Identification Numbers (VINs) of all vehicles driven by that driver, thus aggregating historical manual driving data across vehicles.

[0022] For intelligent driving data, the vehicle system directly aggregates decision data from corresponding algorithms in different scenarios based on the VIN code and the intelligent driving system version number. Then, the vehicle system classifies and processes the data according to four preset sub-scenarios: For the first scenario of green light turning yellow, it extracts road conditions (smooth / congested), vehicle response (stopping / passing), passing speed Vpass, and road speed limit Vlimit; For the second scenario of yellow light turning red, it extracts the remaining time of the yellow light (Tremain) and the speed ratio when the vehicle forcibly passes (the ratio of passing speed Vpass to road speed limit Vlimit); For the third scenario of red light turning green, it extracts the remaining time of the red light (Tstart) and the difference in early start time (the time difference between the moment the red light turns on and the moment of start); For the fourth scenario using intelligent driving, it extracts the intelligent driving status, decision confidence, and human-machine takeover behavior records. Based on the above-mentioned classification data, the manual driving score and intelligent driving risk score are calculated respectively with reference to the classification judgment logic of manual driving and intelligent driving system. Finally, the dynamic risk value is output according to the formula Scorerisk=max(manual driving score, intelligent driving risk score). This value reflects the latest risk status of the current driver or intelligent driving system in real time, and provides core parameters for subsequent intersection collaborative warning.

[0023] The calculation of the manual driving score involves extracting key feature parameters from the first, second, and third scenarios. These parameters include the remaining time of the yellow light (Tremain), the remaining time of the red light (Tstart), the vehicle speed (Vpass), the road speed limit (Vlimit), and the ratio of Vpass to Vlimit, as well as current road conditions (smooth / congested). These parameters are then compared to preset three-level risk thresholds: if a vehicle chooses to proceed through the intersection with a remaining time (Tremain) ≤ 1s in a smooth traffic situation, or forcibly cuts into the queue in a congested traffic situation, or directly runs a red light with a remaining time (Tstart) ≤ 1s... Starting a vehicle that has rushed through a green light within 3 seconds will be considered "dangerous" and will result in a +5 penalty point. If the driver is detected proceeding through the intersection with a yellow light remaining time (Tremain) ≤ 2 seconds, or a red light remaining time (Tstart) ≤ 2 seconds, the driver will be penalized accordingly. If a driver starts moving at a green light within 1 second, they are considered "aggressive" and receive +3 points for that action. If the driver voluntarily stops and waits for the yellow light to pass with a remaining time of >2 seconds, they are considered "conservative" and receive 0 points, meaning no risk score is added.

[0024] In calculating the intelligent driving risk score, in addition to reusing the judgment logic based on physical parameters mentioned above, a special assessment of the algorithm decision-making dimension is also required: When the intelligent driving system is activated, if the system misjudges the traffic light status due to a defect in the perception algorithm, resulting in a violation, the driving behavior will be judged as "dangerous," earning +5 (calibrable) points per instance. If the decision-making and planning module hesitates at the end of the yellow light, leading to a dangerous situation, the driving behavior will be judged as "aggressive," earning +3 (calibrable) points per instance. If the intelligent driving system actively chooses to brake before the yellow light, the driving behavior will be judged as "conservative," earning 0 (calibrable) points per instance. In particular, the vehicle system will conduct a responsibility tracing judgment for human-machine takeover behavior. If the intelligent driving system has issued a clear braking suggestion but the driver forcibly accelerates through, the +5 points will be added to the driver's personal manual driving score. Conversely, if the driver runs a red light due to the intelligent driving system's abnormal exit causing insufficient reaction time, the +5 points will be added to the intelligent driving risk score of the corresponding vehicle model and intelligent driving version. After completing the qualitative judgment and assignment of a single data point, all historical scores of FaceID (for the driver) or VIN + version number (for the intelligent driving system) within the current statistical period will be retrieved. An accumulation algorithm will be used to obtain the manual driving score and the intelligent driving risk score. Finally, a dynamic risk value will be output according to the formula Scorerisk=max(manual driving score, intelligent driving risk score). This ensures that as long as either the driver or the intelligent driving system used exhibits high-risk characteristics, the entire vehicle will be marked as high-risk, thus providing the most conservative and safe decision-making basis for subsequent intersection collaborative defense.

[0025] The risk level is classified as "Dangerous" (+5 points per instance) if any of the following dangerous behaviors are present: continuing to drive in clear traffic with a Tremain ≤ 1s; forcibly cutting in during congested traffic; running a red light; starting before a green light with a Tstart ≤ -3s; intelligent driving system misinterpreting a traffic light violation. The risk level is classified as "Aggressive" (+3 points per instance) if any of the following dangerous behaviors are present: continuing to drive in with a Tremain ≤ 2s; starting before a green light with a Tstart ≤ -1s; intelligent driving system hesitating at the end of a yellow light, leading to danger. The risk level is classified as "Conservative" (+0 points per instance) if any of the following dangerous behaviors are present: actively stopping and waiting with a Tremain > 2s; intelligent driving system actively choosing to brake before the yellow light.

[0026] Step 102: In response to the distance between the vehicle's current location and the location of the intersection it is about to pass through being less than or equal to a preset threshold distance, obtain the risk assessment list corresponding to the current intersection from the cloud based on the current vehicle location and the vehicle's current driving lane.

[0027] The risk assessment list is used to characterize the set of risk attributes of vehicles queuing in each lane at the current intersection.

[0028] In practice, when a vehicle travels along its current lane to within a preset threshold distance (e.g., 50 meters) from the stop line at an intersection, the onboard high-precision positioning module encapsulates the vehicle's current position and current lane (e.g., "Eastbound Straight Lane 2") into a structured query request, which is then uploaded to the cloud server via a 5G / C-V2X communication module. Upon receiving the request, the cloud server first retrieves the high-precision lane topology model of the intersection and, based on a real-time location pool, filters out a list of queuing vehicles located within a preset range before the stop lines of each approach lane and in a low-speed or stationary state. Then, it matches each vehicle in this list with its calculated dynamic risk value and related metadata. The risk assessment list constructed in the cloud not only includes vehicle identification tags (Hash-encrypted VINs or FaceIDs) but also risk value tags, the vehicle's real-time distance from the stop line, vehicle control mode tags (manual driving / intelligent driving), and risk level tags based on points (dangerous / aggressive / conservative). The list is encrypted and then sent to the requesting vehicle, enabling the vehicle to obtain an overview of the risk situation of the entire intersection through cloud interaction before the visual perception has covered the entire intersection. This effectively overcomes the blind spots of single-vehicle sensors in terms of field of view and occlusion.

[0029] Step 103: Perform vehicle consistency matching based on the risk assessment list and the image data of the current intersection collected by the data acquisition equipment to obtain the intersection risk situation map.

[0030] Among them, the intersection risk situation map is used to characterize the dynamic risk distribution of each traffic participant within the current intersection area and their relative spatial location relationship.

[0031] In practice, after receiving the risk assessment list from the cloud, the vehicle initiates a local verification mechanism to eliminate deviations in cloud data caused by latency or lane changes. The vehicle domain controller calls the video stream from the acquisition devices (including front-view and rear-view cameras or 360° panoramic cameras capable of image or video acquisition), extracts the visual features and license plate information of surrounding vehicles through image recognition, and generates a list of surrounding vehicles based on physical space perception. This list of surrounding vehicles is then matched against the vehicle identity tags in the cloud-based risk assessment list: if the match is successful (e.g., the cloud list shows vehicle A 5 meters ahead as "dangerous," and the local camera recognizes that the license plate matches the cloud record), the cloud data is accepted, and an intersection risk situation map is directly constructed based on the cloud data; if the match fails (e.g., the cloud data shows vehicle A ahead as "dangerous," but local recognition shows that the vehicle ahead has left or has been replaced by another vehicle), the cloud-based risk assessment list is dynamically corrected based on the local visual recognition results, invalid data is removed, the vehicle sorting is updated, and a calibrated corrected risk assessment list is generated. Finally, a visualized intersection risk situation map is generated based on the calibrated data. The intersection risk situation map accurately overlays the physical location and risk attributes of vehicles in each lane, providing a high-confidence environmental model for subsequent accurate early warning.

[0032] Step 104: Based on the dynamic risk value and the intersection risk situation map, classify and issue early warnings for the vehicle control mode currently used by the vehicle.

[0033] The vehicle control modes include manual control mode and intelligent driving mode. Manual control mode is where the driver manually controls the vehicle, and control of the vehicle rests with the driver. Intelligent driving mode is where the vehicle's intelligent driving system controls the vehicle, and control of the vehicle rests with the intelligent driving system.

[0034] In practice, based on the generated intersection risk situation map, the vehicle-mounted system comprehensively analyzes the vehicle's dynamic risk value and the surrounding environment, and executes a tiered warning strategy. First, based on collision detection algorithms, target vehicles with potential collision risks and their risk levels are screened from the situation map. For manual control mode, the system focuses on analyzing the spatial positional relationship of the target vehicle relative to the vehicle (e.g., directly in front, to the side, or turning left from the opposite direction). If the intersection risk situation map shows the vehicle ahead is a "dangerous" manually driven vehicle, a red warning message is projected onto the head-up display or central control screen (with simultaneous voice announcement): "The vehicle ahead may run a red light, please keep your distance." If the vehicle ahead is detected as a "conservative" intelligent driving vehicle, the system prompts: "The vehicle ahead is in autonomous driving mode and is expected to stop before the yellow light."

[0035] For the intelligent driving mode, the relative density of high-risk vehicles in the surrounding area is further calculated. If the relative density exceeds a preset threshold (indicating an extremely complex intersection environment), a message is displayed stating "The surrounding environment is complex; driver intervention is recommended," and an audible and visual alarm is triggered. Simultaneously, the intelligent driving system actively reduces the set speed and increases the following distance. If the relative density does not exceed the threshold, normal intelligent driving planning is maintained, and potential risks are highlighted. Furthermore, for high-risk vehicles in the lateral and oncoming lanes (such as vehicles attempting to cut in from the opposite left turn), predictive warnings are provided based on the vehicle's trajectory, such as "There is a risk of oncoming vehicles cutting in; it is recommended to postpone starting." This achieves a shift from passive protection to active defense, improving risk response capabilities.

[0036] In summary, the driving risk warning method provided in this application achieves cross-vehicle aggregated assessment of driver and intelligent driving system risk attributes by collecting cross-vehicle historical manual driving data and cross-vehicle historical intelligent driving data. This constructs a dynamic risk value throughout the entire lifecycle, avoiding the impact of being unable to continuously track driving habits due to being tied to a single vehicle, thus improving the accuracy of risk assessment. Utilizing cloud data interaction, the vehicle can obtain a full-domain risk assessment list when approaching an intersection, overcoming the physical limitations of the onboard sensor's field of view and achieving perception of the macro-level risk situation at the intersection. By introducing image data and consistent matching with the cloud list, the problem of cloud data lag caused by vehicle lane changes and entering / exiting intersections is effectively eliminated, ensuring the authenticity and accuracy of the intersection risk situation map. Combining the vehicle's dynamic risk value with the intersection risk situation map of surrounding vehicles for tiered warnings makes the warning strategy more closely aligned with the actual risk level and the current vehicle control mode, significantly improving the effectiveness of the warning and the user experience.

[0037] In some embodiments, a driving risk assessment is performed based on cross-vehicle historical manual driving data and cross-vehicle historical intelligent driving data to obtain the dynamic risk value of the driver's vehicle at the current moment, including: Based on preset scenario types, cross-vehicle historical manual driving data is classified to obtain scenario driving data. Based on scenario driving data and preset unit time decay coefficient, dangerous behavior is judged to obtain manual driving score. Based on preset scenario types, cross-vehicle historical intelligent driving data is classified to obtain scenario intelligent driving data. Based on the scenario intelligent driving data and the current intelligent driving version, dangerous behaviors are judged to obtain an intelligent driving risk score. The maximum value of the manual driving score and the intelligent driving risk score is determined as the dynamic risk value.

[0038] In practice, the system categorizes and archives driving data according to preset scenario types (Scenario 1, Scenario 2, Scenario 3, and Scenario 4) to form structured scenario-based driving data. Simultaneously, it processes cross-vehicle historical intelligent driving data using the same scenario classification logic to create a scenario-based intelligent driving dataset. For manual driving scoring, the system iterates through and matches scenario data and accumulates points according to preset deduction rules (e.g., +5 points for dangerous driving, +3 points for aggressive driving), and combines this with the initial score for the current scoring period (e.g., a calendar month) to generate an initial manual driving score. For intelligent driving risk scoring, the system not only considers the intelligent driving system's behavior in the scenario data but also introduces a "version correction" mechanism. This involves querying a preset version scoring mapping relationship based on the current intelligent driving version number (e.g., algorithm fluctuations may occur in the early stages of a new version launch, so a correction coefficient is applied), and then overlaying the behavior score with the version correction score to generate an initial intelligent driving score. Finally, the system compares the manual driving score and the intelligent driving risk score, selecting the maximum value as the final output dynamic risk value. This design avoids the one-sidedness of relying solely on human or intelligent driving data, and ensures the conservatism and safety of risk warning by taking the maximum value principle. That is, as long as either human or intelligent driving poses a high risk, the whole vehicle is considered to be in a high-risk state.

[0039] For example, in a manual driving scenario, when User A is driving their own gasoline-powered vehicle (VIN: A123), the vehicle's system records that User A failed to slow down and forcibly crossed the intersection when the yellow light was on and the remaining time (Tremain = 0.8s) was 0.8s (Scenario 2). According to preset rules, User A's driving behavior was deemed "dangerous," resulting in a +5 point deduction. Simultaneously, when User A was driving a rented electric vehicle (VIN: B456), the system recorded that User A failed to slow down and forcibly crossed the intersection when the remaining time (Tstart = 0.8s) was 0.8s. At 2.5 seconds (Scenario 3), the driver started prematurely. User A's driving behavior was judged as "aggressive," earning +3 points for this single instance. Assuming User A has no other violations this month, and according to the attenuation rule (assuming 6 days have passed this month with no new violations, accumulating 6 points attenuation), their initial manual driving score is 5 + 3 = 8 points. After attenuation adjustment, the final manual driving score is 2 points (where 0 points is the lower limit of the score, and the score remains at 0 points after attenuation).

[0040] For the intelligent driving scenario: User A activated the intelligent driving function while driving a vehicle equipped with an intelligent driving system (VIN: C789, intelligent driving version 2.3). During an intersection crossing, the system failed to recognize the traffic light phase change in time when the remaining time (Tremain=1.5s) was too late, causing hesitation in vehicle control and a near-runaway red light. This driving behavior was deemed a "dangerous" decision by the intelligent driving system, earning +5 points. If the initial score for this intelligent driving version is 0 (without a decay mechanism in the intelligent driving score), the final intelligent driving risk score will be +5 points.

[0041] Then, a dynamic risk value is synthesized, extracting the manual driving score and the intelligent driving risk score: the manual driving score is +2 points, and the intelligent driving risk score is +5 points. Comparing the magnitudes of 2 and 5, and based on the principle of taking the maximum value, the current dynamic risk value for user A is determined to be Scorerisk = +5.

[0042] At this point, even if User A's current vehicle is not using the intelligent driving function (i.e., it is currently being driven by a human), and their human driving score has decayed to zero, the system will still determine User A as a "dangerous" risk level driver based on the dynamic risk value Scorerisk=5. When User A approaches an intersection, surrounding vehicles, having learned of User A's high-risk attribute through cloud queries, will maintain a safe distance in advance. Simultaneously, if User A reactivates intelligent driving, the vehicle's system will also, based on this dynamic risk value, prompt the driver to take over or reduce speed when the intersection environment is deemed complex. This dynamic risk value determination mechanism effectively addresses the safety loophole of "mistakes made while driving by a human being being ignored after activating intelligent driving," achieving full lifecycle and full-mode coverage management of driver risk attributes.

[0043] In some embodiments, dangerous behavior is determined based on scenario driving data and a preset unit time decay coefficient to obtain a manual driving score, including: The first set of behaviors is obtained by matching dangerous behaviors with the preset set of dangerous behaviors and the scene driving data. The first behavior score is obtained by accumulating points based on the first behavior set and the preset behavior score mapping relationship. The sum of the first behavior score and the first initial score of the current scoring period is determined as the initial manual driving score. The attenuation score is determined based on the current scoring period and the attenuation coefficient per unit time. The difference between the initial manual driving score and the attenuation score is determined as the manual driving score.

[0044] In practice, the determination process for manual driving scores adopts a dynamic decay mechanism, which is used to reflect the timeliness of driver habits.

[0045] First, after completing dangerous behavior matching and calculating the first behavior score, the first behavior score is added to the initial score of the current scoring cycle (for example, starting from zero o'clock on the 1st day of the current month) (which is positively correlated with the manual driving score at the end of the previous cycle) to obtain the initial manual driving score.

[0046] Wherein, in order to overcome the evaluation discontinuity problem caused by the periodic reset of the scoring system, a scoring inertia mechanism is introduced, that is, the first initial score of the current scoring cycle is dynamically determined. The first initial score is no longer a fixed zero value, but shows a strong positive correlation with the final score of the previous scoring cycle. A historical score archive table for each driver (FaceID) and intelligent driving version (VIN + version number) is stored in the cloud. When a new scoring cycle (such as a natural month) starts, the final manual driving score of the evaluation object in the previous scoring cycle (last month) is retrieved. Then, the first initial score of the current cycle is calculated according to a preset initial score mapping function. The design logic of the initial score mapping function follows the principles of "inertia retention and gradual regression". If the final score of the previous scoring cycle is extremely high (for example, >20 points), it indicates that the driver has deep-rooted dangerous driving habits, and the initial score of the current scoring cycle is set to a relatively high value (for example, the final score × 0.5), which indicates that the driver needs a longer period of compliant driving to reduce the score to a safe level; if the score of the previous scoring cycle is relatively low (for example, <5 points), it indicates that the driving habits are good, and the initial score of the current scoring cycle is set to a relatively low value (even 0) to reward the good performance. The use of the inertia mechanism makes the risk assessment no longer an isolated cycle snapshot, but a continuous dynamic process with a memory function, which effectively avoids the fluke mentality of drivers caused by "score reset to zero" in the previous scoring cycle, and also prevents the "one-vote veto" of historical good performance, reflecting the fairness and scientific nature of the evaluation system.

[0047] Then, the attenuation calculation step is entered. It is monitored in real time whether there is any new dangerous behavior in the current scoring cycle. If there is no new dangerous behavior in the cycle, the time difference from the start time of the cycle to the current time is calculated as the attenuation duration, the duration is multiplied by a preset unit duration attenuation coefficient (for example, 1 point is attenuated per day) to obtain the attenuation score, and the attenuation score is subtracted from the initial manual driving score to obtain the final manual driving score, which reflects the incentive logic of "points decrease for long-term non-violation". If a new dangerous behavior occurs in the cycle, the occurrence time of the earliest target dangerous behavior is located, and only the time difference between the start of the cycle and the occurrence time of the target dangerous behavior is calculated as the effective attenuation duration, based on which the attenuation score is calculated. This means that once a new violation occurs, the subsequent attenuation within this cycle will be blocked, and new points will be superimposed on the original high score, thus avoiding the fluke mentality of drivers that "since points are deducted every day, it does not matter if I run a red light today", and strengthening the disciplinary effect of the scoring.

[0048] In some embodiments, the attenuation score is determined based on the current scoring period and the attenuation coefficient per unit time, including: In response to the absence of new dangerous behaviors within the current scoring period, the time difference between the start time of the current scoring period and the current time is determined as the decay duration, and the product of the decay duration and the decay coefficient per unit time is determined as the decay score. In response to new dangerous behaviors occurring within the current scoring period, the earliest detected target dangerous behavior is identified among the new dangerous behaviors. The time difference between the occurrence time of the target dangerous behavior and the start time of the current scoring period is determined as the decay duration, and the product of the decay duration and the decay coefficient per unit time is determined as the decay score.

[0049] In practice, the attenuation score is used to offset the initial manual driving score. When a new dangerous behavior occurs, the attenuation score calculation is cut off at the moment the first dangerous behavior occurs within the current scoring period. Before performing the attenuation calculation, the risky driving behavior log for the current scoring period must be scanned and judged. In the first scenario, if the scan results show that no new dangerous behaviors are recorded in the current period, the driver is judged to have performed well. In this case, the entire time period between the start time of the current scoring period (e.g., 00:00 on the 1st of this month) and the current time is considered as the attenuation duration, and the attenuation score is determined by multiplying this duration by the attenuation coefficient per unit duration (e.g., 1 point / day). In the second scenario, if the scan finds a new dangerous behavior record within the period, the timestamps of these behaviors are traversed to locate the target dangerous behavior with the earliest timestamp, and the time period between the occurrence time of the target dangerous behavior and the start time of the period is considered as the effective attenuation duration. The truncated calculation method is used because the occurrence of the target dangerous behavior marks the end of the driver's safe driving state. During the time period after the occurrence of this behavior, the driver is no longer eligible to enjoy the points attenuation bonus. This refined time-slicing process more accurately reflects a driver's behavior prior to their most recent violation, while ensuring that the scoring mechanism is not diluted by new violations, thus maintaining the seriousness and accuracy of risk assessment.

[0050] In some embodiments, dangerous behaviors are determined based on scene-specific intelligent driving data and the current intelligent driving version to obtain an intelligent driving risk score, including: A second set of behaviors is obtained by matching dangerous behaviors with a preset set of dangerous behaviors and scene-based intelligent driving data; The second behavior score is obtained by accumulating points based on the second behavior set and the preset behavior score mapping relationship. The sum of the second behavior score and the second initial score of the current intelligent driving version is determined as the initial intelligent driving score. The version correction score corresponding to the current intelligent driving version is determined based on the preset version score mapping relationship, and the sum of the initial intelligent driving score and the version correction score is determined as the intelligent driving risk score.

[0051] In practical implementation, a scoring model distinct from human driving was constructed to address the unique characteristics of intelligent driving systems. First, the collected cross-vehicle historical intelligent driving data is categorized by scenario, forming a scenario-based intelligent driving dataset. This scenario data is then matched with a pre-defined set of dangerous behaviors to identify poor decision-making by the intelligent driving system in specific scenarios (such as hesitation at the end of a yellow light, misjudging traffic lights, etc.). Points are accumulated based on the behavior scoring mapping relationship to obtain a second behavior score. Considering that the risks of intelligent driving systems stem more from the algorithm itself than from driver habits, a version correction mechanism is introduced. The second behavior score is added to the second initial score of the current intelligent driving version (usually set to 0, or given an initial rating based on industry evaluations) to obtain the initial intelligent driving score.

[0052] Then, the preset version rating mapping table is queried. The rating mapping table records the historical performance data or expected stability indicators of different intelligent driving versions. The version correction score corresponding to the current intelligent driving version is extracted (for example, if the new version fixes a known bug, it can be given a negative correction, i.e., a deduction of points; if the old version has hidden dangers, it can be given a positive correction, i.e., a bonus of points).

[0053] Finally, the initial intelligent driving score and the version correction score are summed to obtain the final intelligent driving risk score. The version correction mechanism ensures that the intelligent driving risk score not only reflects a single violation but also relates to the inherent attributes of the software version, providing a quantitative basis for optimizing the algorithm and assessing intelligent driving safety.

[0054] In some embodiments, based on the current vehicle location and the vehicle's current lane, a risk assessment list corresponding to the current intersection is obtained from the cloud, including: Based on the current vehicle location and the current lane, a query request is sent to the cloud so that the cloud can determine the lane queuing list corresponding to the current intersection. For each vehicle in the lane queuing list, a vehicle identity tag, risk value tag, stop line spacing tag, vehicle control mode tag, and risk level tag are added to obtain a risk assessment list sent to the vehicle.

[0055] In practice, when a vehicle approaches an intersection, its onboard communication unit sends a query request to the cloud server containing the vehicle's current location (latitude and longitude coordinates), current lane (lane ID), direction of travel, and vehicle identification information. After parsing the query request, the cloud server matches the corresponding intersection geofence based on the coordinates and then retrieves the latest location reports of all online vehicles within that geofence. The cloud then groups and sorts these vehicles according to their lane IDs, constructing a lane queuing list reflecting the actual queuing order of vehicles in each lane. Next, for each vehicle in the lane queuing list, the cloud performs data enrichment operations, querying the vehicle's identification tag (FaceID or VIN), querying its latest dynamic risk value, calculating its real-time distance from the stop line, reading its current vehicle control mode (human-driven / intelligent-driven), and determining its risk level (dangerous / aggressive / conservative) based on the score range and dynamic risk value. This tag data is packaged into a risk assessment list distributed to the vehicle. In particular, to protect privacy, the cloud can perform irreversible hash encryption on license plates or Face IDs before sending data, retaining only the necessary collision warning information, thus achieving multi-vehicle collaborative perception while ensuring the privacy and security of user data.

[0056] In some embodiments, vehicle consistency matching is performed based on the risk assessment list and the image data of the current intersection collected by the acquisition device to obtain an intersection risk situation map, including: Vehicle identification is performed based on image data to obtain a list of surrounding vehicles; The vehicle identity is matched for consistency based on the vehicle identity tags in the surrounding vehicle list and the risk assessment list. Upon successful consistency matching, an intersection risk situation map is constructed based on the risk assessment list. In response to a failure to match a consistent risk profile, the risk assessment list is revised based on the list of surrounding vehicles to obtain a revised risk assessment list, and an intersection risk situation map is constructed based on the revised risk assessment list.

[0057] In practice, to ensure alignment between virtual and real data, the vehicle's visual perception module continuously processes image data, detecting vehicle outlines, license plate characters, and vehicle feature points in the images, generating a list of surrounding vehicles including license plate numbers, relative distances, and azimuth angles. The license plate information in this list is then compared one by one with the vehicle identity tags (decrypted license plates or IDs) carried in the cloud-based risk assessment list. If the comparison results match (e.g., the surrounding list identifies a license plate "XAxxxxx" located 5 meters in front of the vehicle, and the corresponding vehicle ID in the cloud list is also located 5 meters in front), the match is considered successful, the cloud data is trusted, and the risk values ​​in the cloud list are directly used to construct an intersection risk situation map.

[0058] If the comparison fails (e.g., the cloud list shows the vehicle ahead as high-risk, but local visual recognition detects the vehicle is no longer in sight or the license plate does not match), it is determined that the cloud data is invalid due to the vehicle changing lanes, leaving the lane, or network latency. In this case, a correction process is initiated. Based on the surrounding vehicle list obtained from local visual recognition, vehicles not in the cloud list are removed, existing vehicles are retained, and newly appearing vehicles (if they can be recognized locally but are not in the cloud list) are marked, generating a corrected risk assessment list. Finally, a risk situation map of the intersection is constructed based on the corrected list, ensuring the physical authenticity and timeliness of the system's decision-making basis.

[0059] The intersection risk situation map refers to a structured data model and visualization logical topology generated by an on-board domain controller or cloud server, representing the dynamic risk distribution and relative spatial relationships of various traffic participants within a specific intersection area. This situation map is not a simple geographical map, but a multi-dimensional spatiotemporal dataset that integrates static lane topology, dynamic vehicle status, and quantified risk attributes.

[0060] In some embodiments, a graded early warning is provided for the vehicle control mode currently used by the vehicle based on dynamic risk values ​​and intersection risk situation maps, including: Based on the dynamic risk value and the intersection risk situation map, the target vehicles with collision risk are identified, and the target risk level of the target vehicles is determined based on the intersection risk situation map. In response to the vehicle control mode being manual control mode, the relative positional relationship is determined based on the target vehicle's position and the current vehicle's position, and a safety warning is issued based on the relative positional relationship and risk level. In response to the vehicle control mode being set to intelligent driving mode, dangerous vehicles with a target risk level greater than or equal to a preset threshold level are identified among the target vehicles, and the vehicle density of dangerous vehicles is determined based on the intersection risk situation map. In response to vehicle density being greater than or equal to a preset threshold density, a manual intervention warning will be issued; In response to a vehicle density that is less than a preset threshold, the intersection speed is reduced while issuing a safety warning.

[0061] In practice, a fusion analysis is performed based on dynamic risk values ​​(reflecting the vehicle's attributes) and intersection risk situation maps (reflecting environmental attributes) to first identify target vehicles with collision risks on the predicted trajectory. For manual control mode, the focus is on defense against vehicles in front and to the sides. Based on the spatial position of the target vehicle relative to the vehicle (e.g., cutting in from directly in front or to the left rear) and the target vehicle's target risk level, differentiated head-up display text prompts or voice alerts are generated. For example, when a high-risk vehicle is detected in front, a prompt to maintain distance is given; or when an oncoming vehicle is detected making a left turn and cutting in front, a prompt to pause the start is given.

[0062] For the intelligent driving mode, the focus is on system safety redundancy. After filtering out target vehicles with risk levels higher than a threshold (such as "dangerous"), the system calculates the distribution density of these dangerous vehicles in the surrounding environment (such as the number of dangerous vehicles per unit radius). If the vehicle density exceeds a preset threshold, it indicates that the complexity of the current intersection environment exceeds the preset safety boundary of the intelligent driving system. The system will then forcibly trigger a "manual intervention warning," prompting the driver to take over control through a head-up display, seatbelt pretensioning, and audible alarm. If the density does not exceed the threshold, while maintaining the intelligent driving state, the system will appropriately reduce the intersection speed (such as reducing the speed limit by 10%-20%) and activate more sensitive collision avoidance monitoring. This adaptive strategy based on the relative density of dangerous vehicles effectively balances the traffic efficiency of autonomous driving with driving safety in extreme environments.

[0063] It should be noted that the method in this embodiment can be executed by a single device, such as a computer or server. The method can also be applied in a distributed scenario, where multiple devices cooperate to complete the task. In such a distributed scenario, one of these devices may execute only one or more steps of the method in this embodiment, and the multiple devices will interact with each other to complete the method described.

[0064] It should be noted that the above description describes some embodiments of this application. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in a different order than that shown in the above embodiments and still achieve the desired result. Furthermore, the processes depicted in the drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0065] Based on the same inventive concept, corresponding to any of the above embodiments, this application also provides a driving risk warning device.

[0066] refer to Figure 2 The driving risk warning device includes: The risk dynamic assessment module 10 is configured to: conduct driving risk assessment based on cross-vehicle historical manual driving data and cross-vehicle historical intelligent driving data to obtain the dynamic risk value of the driver's vehicle at the current moment; wherein, the cross-vehicle historical manual driving data is the historical manual driving behavior data associated with the driver's identity and covering different vehicles; the cross-vehicle historical intelligent driving data is the historical intelligent driving decision data associated with the vehicle identification code and intelligent driving version of each vehicle. The risk list acquisition module 20 is configured to: in response to the interval between the current vehicle position and the intersection position to be passed being less than or equal to a preset threshold distance, acquire the risk assessment list corresponding to the current intersection from the cloud based on the current vehicle position and the current driving lane of the vehicle; the risk assessment list is used to characterize the risk attribute set of vehicles queuing in each lane of the current intersection. The risk perception module 30 is configured to: perform vehicle consistency matching based on the risk assessment list and the image data of the current intersection collected by the acquisition device to obtain an intersection risk situation map; the intersection risk situation map is used to characterize the dynamic risk distribution and relative spatial position relationship of each traffic participant within the current intersection range; The graded early warning module 40 is configured to: provide graded early warnings based on dynamic risk values ​​and intersection risk situation maps for the vehicle control mode currently used by the vehicle; the vehicle control mode includes manual control mode or intelligent driving mode.

[0067] In one embodiment, the risk dynamic assessment module 10 is further configured to: Based on preset scenario types, cross-vehicle historical manual driving data is classified to obtain scenario driving data. Based on scenario driving data and preset unit time decay coefficient, dangerous behavior is judged to obtain manual driving score. Based on preset scenario types, cross-vehicle historical intelligent driving data is classified to obtain scenario intelligent driving data. Based on the scenario intelligent driving data and the current intelligent driving version, dangerous behaviors are judged to obtain an intelligent driving risk score. The maximum value of the manual driving score and the intelligent driving risk score is determined as the dynamic risk value.

[0068] In one embodiment, the risk dynamic assessment module 10 is further configured to: The first set of behaviors is obtained by matching dangerous behaviors with the preset set of dangerous behaviors and the scene driving data. The first behavior score is obtained by accumulating points based on the first behavior set and the preset behavior score mapping relationship. The sum of the first behavior score and the first initial score of the current scoring period is determined as the initial manual driving score. The attenuation score is determined based on the current scoring period and the attenuation coefficient per unit time. The difference between the initial manual driving score and the attenuation score is determined as the manual driving score.

[0069] In one embodiment, the risk dynamic assessment module 10 is further configured to: In response to the absence of new dangerous behaviors within the current scoring period, the time difference between the start time of the current scoring period and the current time is determined as the decay duration, and the product of the decay duration and the decay coefficient per unit time is determined as the decay score. In response to new dangerous behaviors occurring within the current scoring period, the earliest detected target dangerous behavior is identified among the new dangerous behaviors. The time difference between the occurrence time of the target dangerous behavior and the start time of the current scoring period is determined as the decay duration, and the product of the decay duration and the decay coefficient per unit time is determined as the decay score.

[0070] In one embodiment, the risk dynamic assessment module 10 is further configured to: A second set of behaviors is obtained by matching dangerous behaviors with a preset set of dangerous behaviors and scene-based intelligent driving data; The second behavior score is obtained by accumulating points based on the second behavior set and the preset behavior score mapping relationship. The sum of the second behavior score and the second initial score of the current intelligent driving version is determined as the initial intelligent driving score. The version correction score corresponding to the current intelligent driving version is determined based on the preset version score mapping relationship, and the sum of the initial intelligent driving score and the version correction score is determined as the intelligent driving risk score.

[0071] In one embodiment, the risk list acquisition module 20 is further configured to: Based on the current vehicle location and the current lane, a query request is sent to the cloud so that the cloud can determine the lane queuing list corresponding to the current intersection. For each vehicle in the lane queuing list, a vehicle identity tag, risk value tag, stop line spacing tag, vehicle control mode tag, and risk level tag are added to obtain a risk assessment list sent to the vehicle.

[0072] In one embodiment, the risk perception module 30 is further configured to: Vehicle identification is performed based on image data to obtain a list of surrounding vehicles; The vehicle identity is matched for consistency based on the vehicle identity tags in the surrounding vehicle list and the risk assessment list. Upon successful consistency matching, an intersection risk situation map is constructed based on the risk assessment list. In response to a failure to match a consistent risk profile, the risk assessment list is revised based on the list of surrounding vehicles to obtain a revised risk assessment list, and an intersection risk situation map is constructed based on the revised risk assessment list.

[0073] In one embodiment, the graded early warning module 40 is further configured to: Based on the dynamic risk value and the intersection risk situation map, the target vehicles with collision risk are identified, and the target risk level of the target vehicles is determined based on the intersection risk situation map. In response to the vehicle control mode being manual control mode, the relative positional relationship is determined based on the target vehicle's position and the current vehicle's position, and a safety warning is issued based on the relative positional relationship and risk level. In response to the vehicle control mode being set to intelligent driving mode, dangerous vehicles with a target risk level greater than or equal to a preset threshold level are identified among the target vehicles, and the vehicle density of dangerous vehicles is determined based on the intersection risk situation map. In response to vehicle density being greater than or equal to a preset threshold density, a manual intervention warning will be issued; In response to a vehicle density that is less than a preset threshold, the intersection speed is reduced while issuing a safety warning.

[0074] For ease of description, the above devices are described in terms of function, divided into various modules. Of course, in implementing this application, the functions of each module can be implemented in one or more software and / or hardware.

[0075] The apparatus of the above embodiments is used to implement the corresponding driving risk warning method in any of the foregoing embodiments, and has the beneficial effects of the corresponding method embodiments, which will not be repeated here.

[0076] Based on the same inventive concept, corresponding to the methods of any of the above embodiments, this application also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the driving risk warning method described in any of the above embodiments.

[0077] Figure 3 This embodiment illustrates a more specific hardware structure of an electronic device. The device may include a processor 1010, a memory 1020, an input / output interface 1030, a communication interface 1040, and a bus 1050. The processor 1010, memory 1020, input / output interface 1030, and communication interface 1040 are interconnected internally via the bus 1050.

[0078] The processor 1010 can be implemented using a general-purpose CPU (Central Processing Unit), microprocessor, application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of this specification.

[0079] The memory 1020 can be implemented in the form of ROM (Read Only Memory), RAM (Random Access Memory), static storage device, dynamic storage device, etc. The memory 1020 can store the operating system and other applications. When the technical solutions provided in the embodiments of this specification are implemented by software or firmware, the relevant program code is stored in the memory 1020 and is called and executed by the processor 1010.

[0080] The input / output interface 1030 is used to connect input / output modules to realize information input and output. Input / output modules can be configured as components within the device (not shown in the figure) or externally connected to the device to provide corresponding functions. Input devices may include keyboards, mice, touchscreens, microphones, various sensors, etc., while output devices may include displays, speakers, vibrators, indicator lights, etc.

[0081] The communication interface 1040 is used to connect a communication module (not shown in the figure) to enable communication between this device and other devices. The communication module can communicate via wired means (such as USB, Ethernet cable, etc.) or wireless means (such as mobile network, WIFI, Bluetooth, etc.).

[0082] Bus 1050 includes a pathway for transmitting information between various components of the device, such as processor 1010, memory 1020, input / output interface 1030, and communication interface 1040.

[0083] It should be noted that although the above-described device only shows the processor 1010, memory 1020, input / output interface 1030, communication interface 1040, and bus 1050, in specific implementations, the device may also include other components necessary for normal operation. Furthermore, those skilled in the art will understand that the above-described device may only include the components necessary for implementing the embodiments of this specification, and not necessarily all the components shown in the figures.

[0084] The electronic devices described above are used to implement the corresponding driving risk warning methods in any of the foregoing embodiments, and have the beneficial effects of the corresponding method embodiments, which will not be repeated here.

[0085] Based on the same inventive concept, corresponding to the methods of any of the above embodiments, this application also provides a non-transitory computer-readable storage medium storing computer instructions for causing the computer to execute the driving risk warning method as described in any of the above embodiments.

[0086] The computer-readable medium of this embodiment includes permanent and non-permanent, removable and non-removable media, and information storage can be implemented by any method or technology. Information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transfer medium that can be used to store information accessible by a computing device.

[0087] The computer instructions stored in the storage medium of the above embodiments are used to cause the computer to execute the driving risk warning method as described in any of the above embodiments, and have the beneficial effects of the corresponding method embodiments, which will not be repeated here.

[0088] Based on the same inventive concept, corresponding to any of the above embodiments, this application also provides a vehicle, including the electronic equipment or driving risk warning device of the above embodiments, and executes the driving risk warning method as described in any of the above embodiments through the electronic equipment or driving risk warning device of the above embodiments, and has the beneficial effects of the corresponding method embodiments, which will not be repeated here.

[0089] It is understood that before using the technical solutions of the various embodiments in this application, users will be informed of the type, scope of use, and usage scenarios of the personal information involved in an appropriate manner, and user authorization will be obtained.

[0090] For example, upon receiving a user's active request, a prompt message is sent to the user to explicitly inform them that the requested operation will require the acquisition and use of the user's personal information. This allows the user to independently choose, based on the prompt message, whether to provide personal information to the software or hardware such as electronic devices, applications, servers, or storage media performing the operations described in this application.

[0091] As an optional but not limited implementation, in response to a user's active request, sending a prompt message to the user can be done via a pop-up window, where the prompt message can be presented in text format. Furthermore, the pop-up window can also include a selection control allowing the user to choose "agree" or "disagree" to provide personal information to the electronic device.

[0092] It is understood that the above notification and user authorization process is merely illustrative and does not limit the implementation of this application. Other methods that comply with relevant laws and regulations may also be applied to the implementation of this application.

[0093] Those skilled in the art should understand that the discussion of any of the above embodiments is merely exemplary and is not intended to imply that the scope of this application is limited to these examples; under the concept of this application, the technical features of the above embodiments or different embodiments can also be combined, the steps can be implemented in any order, and there are many other variations of different aspects of the embodiments of this application as described above, which are not provided in detail for the sake of brevity.

[0094] Additionally, to simplify the description and discussion, and to avoid obscuring the embodiments of this application, the well-known power / ground connections to integrated circuit (IC) chips and other components may or may not be shown in the provided drawings. Furthermore, the apparatus may be shown in block diagram form to avoid obscuring the embodiments of this application, and this also takes into account the fact that the details of the implementation of these block diagram apparatuses are highly dependent on the platform on which the embodiments of this application will be implemented (i.e., these details should be fully understood by those skilled in the art). While specific details (e.g., circuits) have been set forth to describe exemplary embodiments of this application, it will be apparent to those skilled in the art that the embodiments of this application can be implemented without these specific details or with variations thereof. Therefore, these descriptions should be considered illustrative rather than restrictive.

[0095] Although this application has been described in conjunction with specific embodiments thereof, many substitutions, modifications, and variations of these embodiments will be apparent to those skilled in the art from the foregoing description. For example, other memory architectures (e.g., dynamic RAM (DRAM)) may be used with the embodiments discussed.

[0096] The embodiments of this application are intended to cover all such substitutions, modifications, and variations that fall within the broad scope of the claims of this application. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the embodiments of this application should be included within the protection scope of this application.

Claims

1. A method for early warning of driving risks, characterized in that, include: Driving risk assessment is performed based on cross-vehicle historical manual driving data and cross-vehicle historical intelligent driving data to obtain the dynamic risk value of the driver's vehicle at the current moment; wherein, the cross-vehicle historical manual driving data is historical manual driving behavior data associated with the driver's identity and covering different vehicles; the cross-vehicle historical intelligent driving data is historical intelligent driving decision data associated with the vehicle identification code and intelligent driving version of each vehicle; In response to the fact that the distance between the vehicle's current position and the intersection it is about to pass through is less than or equal to a preset threshold distance, a risk assessment list corresponding to the current intersection is obtained from the cloud based on the current vehicle position and the vehicle's current driving lane; the risk assessment list is used to characterize the risk attribute set of vehicles queuing in each lane of the current intersection. Based on the risk assessment list and the image data of the current intersection collected by the acquisition device, vehicle consistency matching is performed to obtain an intersection risk situation map; the intersection risk situation map is used to characterize the dynamic risk distribution and relative spatial position relationship of each traffic participant within the current intersection range; The vehicle control mode currently used by the vehicle is classified and warned according to the dynamic risk value and the intersection risk situation map; the vehicle control mode includes manual control mode or intelligent driving mode.

2. The driving risk warning method according to claim 1, characterized in that, The driving risk assessment based on cross-vehicle historical manual driving data and cross-vehicle historical intelligent driving data yields the dynamic risk value of the driver's vehicle at the current moment, including: The cross-vehicle historical manual driving data is classified according to the preset scenario type to obtain scenario driving data. Dangerous behavior is judged based on the scenario driving data and the preset unit time decay coefficient to obtain a manual driving score. The cross-vehicle historical intelligent driving data is classified according to the preset scenario type to obtain scenario intelligent driving data. Dangerous behaviors are judged based on the scenario intelligent driving data and the current intelligent driving version to obtain an intelligent driving risk score. The maximum value of the manual driving score and the intelligent driving risk score is determined as the dynamic risk value.

3. The driving risk warning method according to claim 2, characterized in that, The step of determining dangerous behavior based on the scenario driving data and a preset unit time decay coefficient to obtain a manual driving score includes: A first set of behaviors is obtained by matching dangerous behaviors with the scenario driving data based on a preset set of dangerous behaviors. The first behavior score is obtained by accumulating points based on the first behavior set and the preset behavior score mapping relationship. The sum of the first behavior score and the first initial score of the current scoring period is determined as the initial manual driving score. The attenuation score is determined based on the current scoring period and the unit duration attenuation coefficient, and the difference between the initial manual driving score and the attenuation score is determined as the manual driving score.

4. The driving risk warning method according to claim 3, characterized in that, The step of determining the attenuation score based on the current scoring period and the unit duration attenuation coefficient includes: In response to the absence of new dangerous behaviors within the current scoring period, the time difference between the start time of the current scoring period and the current time is determined as the decay duration, and the product of the decay duration and the unit duration decay coefficient is determined as the decay score; In response to a new dangerous behavior occurring within the current scoring period, the earliest detected target dangerous behavior is identified from the new dangerous behavior. The time difference between the occurrence time of the target dangerous behavior and the start time of the current scoring period is determined as the decay duration, and the product of the decay duration and the unit duration decay coefficient is determined as the decay score.

5. The method for early warning of driving risks according to claim 2, characterized in that, The step of determining dangerous behaviors based on the scenario's intelligent driving data and the current intelligent driving version to obtain an intelligent driving risk score includes: A second set of behaviors is obtained by matching dangerous behaviors with the scenario-based intelligent driving data based on a preset set of dangerous behaviors. The second behavior score is obtained by accumulating points based on the second behavior set and the preset behavior score mapping relationship. The sum of the second behavior score and the second initial score of the current intelligent driving version is determined as the initial intelligent driving score. The version correction score corresponding to the current intelligent driving version is determined according to the preset version score mapping relationship, and the sum of the initial intelligent driving score and the version correction score is determined as the intelligent driving risk score.

6. The method for early warning of driving risks according to claim 1, characterized in that, The step of obtaining a risk assessment list corresponding to the current intersection from the cloud based on the current vehicle location and the vehicle's current lane includes: Based on the current vehicle location and the current lane, a query request is sent to the cloud so that the cloud can determine the lane queuing list corresponding to the current intersection, and add a vehicle identity tag, risk value tag, stop line spacing tag, vehicle control mode tag and risk level tag to each vehicle in the lane queuing list to obtain the risk assessment list sent to the vehicle.

7. The method for early warning of driving risks according to claim 1, characterized in that, The step of performing vehicle consistency matching based on the risk assessment list and the image data of the current intersection collected by the acquisition device to obtain an intersection risk situation map includes: Vehicle identification is performed based on the image data to obtain a list of surrounding vehicles; The vehicle identity is matched for consistency based on the vehicle identity tags in the surrounding vehicle list and the risk assessment list. Upon successful consistency matching, the intersection risk situation map is constructed based on the risk assessment list. In response to a consistency matching failure, the risk assessment list is revised based on the surrounding vehicle list to obtain a revised risk assessment list, and the intersection risk situation map is constructed based on the revised risk assessment list.

8. The method for early warning of driving risks according to claim 1, characterized in that, The step of providing graded early warnings based on the dynamic risk value and the intersection risk situation map for the vehicle's current vehicle control mode includes: Based on the dynamic risk value and the intersection risk situation map, target vehicles with collision risk are identified, and the target risk level of the target vehicles is determined based on the intersection risk situation map. In response to the vehicle control mode being manual control mode, a relative positional relationship is determined based on the target vehicle position and the current vehicle position, and a safety warning is issued based on the relative positional relationship and the risk level. In response to the vehicle control mode being set to intelligent driving mode, dangerous vehicles with a target risk level greater than or equal to a preset threshold level are identified among the target vehicles, and the vehicle density of the dangerous vehicles is determined based on the intersection risk situation map. In response to the vehicle density being greater than or equal to a preset threshold density, a manual intervention warning is issued; In response to the vehicle density being less than a preset threshold density, the intersection speed is reduced while issuing a safety warning.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the program, it implements the method as described in any one of claims 1 to 8.

10. A vehicle, characterized in that, Including the electronic device as described in claim 9.

Images