Vehicle real-time risk degree evaluation method and device

Abstract

The invention provides a vehicle real-time risk degree evaluation method and device, and is applied to the technical field of automatic driving. The method comprises the following steps: acquiring operation data of a vehicle and environment data of a driving environment of the vehicle; and determining a risk dimension index of the vehicle in the current driving environment according to the operation data and the environment data. And dynamically adjusting the weight coefficient of the risk dimension index according to the risk dimension index of the vehicle and the historical risk dimension index of the vehicle. And performing calculation based on the dynamically adjusted weight coefficient and the risk dimension index to obtain a real-time risk index of the vehicle, thereby solving the problems of single evaluation dimension, poor scene adaptability and lack of dynamic adjustment capability in a traditional vehicle real-time risk evaluation method.

Description

Technical Field

[0001] This application relates to the field of autonomous driving technology, and in particular to a method and apparatus for real-time vehicle hazard assessment. Background Technology

[0002] With the rapid development of intelligent vehicles, the performance evaluation of autonomous vehicles is a core link supporting technology research and development verification and commercialization. However, related technologies still face severe challenges in terms of technological evolution and adaptation to complex scenarios.

[0003] As autonomous driving technology evolves from basic control functions to advanced decision-making intelligence, the industry's performance evaluation requirements have shifted from single-function stability to a holistic measurement of the system's overall performance. However, traditional evaluation methods generally suffer from the problem of deep coupling between indicator construction and scenarios. Evaluation results heavily depend on specific test scenarios, making it difficult to maintain stability and comparability across different road types, traffic flow states, and operating environments. For example, evaluation indicators built based on fixed scenarios often exhibit drastic fluctuations in scores when encountering unpredictable traffic participant behaviors or complex weather conditions, failing to accurately reflect system performance.

[0004] On the other hand, most of these evaluation methods rely on preset thresholds or rule combinations, lacking a continuous and adaptive characterization of the system's dynamic operating state. While such methods perform well on specific test sets, they struggle to depict the system's true performance in long-tail scenarios and are unable to proactively assess potential risks. Furthermore, existing methods often focus on characterizing a single performance aspect, lacking a comprehensive evaluation framework capable of integrating multi-source information and possessing time-series modeling capabilities, leading to biased or one-sided evaluation results. Summary of the Invention

[0005] The purpose of this application is to provide a method and device for real-time vehicle hazard assessment, which solves the problems of traditional real-time vehicle hazard assessment methods, such as single assessment dimensions, poor scenario adaptability, and lack of dynamic adjustment capabilities.

[0006] Firstly, embodiments of this application provide a method for real-time vehicle hazard assessment. The method includes: acquiring vehicle operating data and environmental data of the vehicle's driving environment; determining risk dimension indicators for the vehicle in the current driving environment based on the operating data and environmental data; dynamically adjusting the weight coefficients of the risk dimension indicators based on the vehicle's risk dimension indicators and historical risk dimension indicators; and calculating the vehicle's real-time hazard index based on the dynamically adjusted weight coefficients and the risk dimension indicators.

[0007] This application provides a method for real-time vehicle hazard assessment. By acquiring vehicle operational data and environmental data of the vehicle's driving environment, a risk dimension index for the vehicle in the current driving environment is determined based on this data. The weight of this risk dimension index is then dynamically adjusted using the vehicle's risk dimension index and its historical risk dimension index. This allows the weight coefficient to be dynamically adjusted according to the actual driving scenario, improving the adaptability and accuracy of the vehicle's real-time hazard assessment under different road types, traffic conditions, and weather conditions. Finally, based on the dynamically adjusted weight coefficients and risk dimension index, a real-time hazard index for the vehicle in the current environment is calculated to assist in the autonomous driving control of the vehicle, providing a more reliable control basis for the decision-making and control of autonomous driving.

[0008] One possible implementation involves using environmental data including at least one traffic participant surrounding the vehicle. When the risk dimension indicator is an instantaneous safety risk indicator, the risk dimension indicator for the vehicle in its current driving environment is determined based on operational data and environmental data. This includes: for any target traffic participant among all traffic participants surrounding the vehicle, determining the repulsive potential field between the vehicle and the target traffic participant based on the vehicle's current position in the operational data and the target traffic participant's current position in the environmental data. The repulsive potential field is then corrected using spatial risk sensitivity and temporal risk sensitivity to obtain the target instantaneous safety risk indicator between the vehicle and the target traffic participant. The maximum value among the target instantaneous safety risk indicators for all target traffic participants surrounding the vehicle is determined as the instantaneous safety risk indicator.

[0009] One possible implementation, where the risk dimension indicator is a traffic efficiency indicator, involves determining the vehicle's risk dimension indicator in the current driving environment based on operational and environmental data. This includes: determining the vehicle's average speed based on the vehicle's speed at each moment within a preset time window from the operational data; obtaining the average speed of the traffic flow within the preset time window from the environmental data; and determining the traffic efficiency indicator based on the vehicle's average speed, the average traffic flow speed, and the weather impact coefficient corresponding to the vehicle's environment from the environmental data.

[0010] One possible implementation, where the risk dimension index is the ride comfort index, involves determining the vehicle's risk dimension index in the current driving environment based on operational and environmental data. This includes: determining the instantaneous acceleration of the vehicle along multiple preset axes in the vehicle's coordinate system based on the vehicle's acceleration in the operational data; weighting and filtering the target instantaneous acceleration for any one of the preset axes to obtain the weighted and filtered target instantaneous acceleration; calculating the root mean square value of the weighted and filtered target instantaneous acceleration; and weighted and fusing the root mean square values ​​of the target instantaneous accelerations along all preset axes to obtain the ride comfort index.

[0011] One possible implementation involves dynamically adjusting the weighting coefficients of risk dimension indicators based on the vehicle's risk dimension indicators and historical risk dimension indicators. This includes: determining the indicator error corresponding to each risk dimension indicator based on the risk dimension indicator and its corresponding safety threshold; determining a correction value for the indicator error based on the indicator error and the historical indicator error corresponding to the vehicle's historical risk dimension indicators; determining a baseline value for each risk dimension indicator based on all risk dimension indicators within a preset time window; and correcting the baseline value using the correction value to obtain the dynamically adjusted weighting coefficients.

[0012] One possible implementation, the vehicle real-time hazard assessment method provided in this application embodiment, further includes: normalizing all dynamically adjusted weight coefficients to obtain the dynamically adjusted weight coefficients corresponding to each risk dimension index in the risk dimension index.

[0013] One possible implementation of the vehicle real-time hazard assessment method provided in this application embodiment further includes: determining whether the vehicle poses a driving hazard based on a real-time hazard index. If a driving hazard is determined, a vehicle control strategy is determined based on the real-time hazard index to control the vehicle.

[0014] One possible implementation involves determining a vehicle control strategy based on real-time hazard indicators, including: maintaining the vehicle's autonomous driving state and recording data when the real-time hazard indicators are determined to be in a low-risk range; and / or generating a takeover warning and pre-tightening the seatbelts when the real-time hazard indicators are determined to be in a medium-risk range to remind the driver to take over the vehicle; and / or triggering the vehicle's emergency braking function or controlling the vehicle to perform a pullover operation when the real-time hazard indicators are determined to be in a high-risk range, and uploading accident data.

[0015] Secondly, embodiments of this application provide a vehicle real-time hazard assessment device, which includes: an acquisition module, a determination module, an adjustment module, and a calculation module.

[0016] The acquisition module is used to acquire vehicle operating data and environmental data of the vehicle's driving environment.

[0017] The determination module is used to determine the risk dimension indicators of the vehicle in the current driving environment based on operational data and environmental data.

[0018] The adjustment module is used to dynamically adjust the weight coefficients of the risk dimension indicators based on the vehicle's risk dimension indicators and the vehicle's historical risk dimension indicators.

[0019] The calculation module is used to calculate the real-time hazard index of the vehicle based on the dynamically adjusted weight coefficients and risk dimension indicators.

[0020] One possible implementation involves using environmental data including at least one traffic participant surrounding the vehicle. When the risk dimension indicator is an instantaneous safety risk indicator, the determination module, when determining the vehicle's risk dimension indicator in the current driving environment based on operational data and environmental data, specifically performs the following: For any target traffic participant among all traffic participants surrounding the vehicle, based on the vehicle's current position in the operational data and the target traffic participant's current position in the environmental data, determine the repulsive potential field between the vehicle and the target traffic participant. The repulsive potential field is then corrected using spatial risk sensitivity and temporal risk sensitivity to obtain the target instantaneous safety risk indicator between the vehicle and the target traffic participant. The maximum value among the target instantaneous safety risk indicators of all target traffic participants surrounding the vehicle is determined as the instantaneous safety risk indicator.

[0021] One possible implementation, where the risk dimension indicator is a traffic efficiency indicator, involves the following steps when the module determines the risk dimension indicator of a vehicle in the current driving environment based on operational and environmental data: First, it determines the vehicle's average speed based on the vehicle's speed at each moment within a preset time window from the operational data. Second, it obtains the average speed of the traffic flow within the preset time window from the environmental data. Finally, it determines the traffic efficiency indicator based on the vehicle's average speed, the average traffic flow speed, and the weather influence coefficient corresponding to the vehicle's environment from the environmental data.

[0022] One possible implementation, where the risk dimension indicator is a ride comfort indicator, involves the following steps when the determination module, based on operational and environmental data, determines the vehicle's risk dimension indicator in the current driving environment: First, based on the vehicle's acceleration in the operational data, determine the instantaneous acceleration of the vehicle along multiple preset axes in the vehicle's coordinate system. Then, for any one of these preset axes, perform a weighted filtering on the target instantaneous acceleration to obtain a weighted filtered target instantaneous acceleration. Next, calculate the root mean square (RMS) value of the weighted filtered target instantaneous acceleration. Finally, weightedly fuse the RMS values ​​of the target instantaneous accelerations along all preset axes to obtain the ride comfort indicator.

[0023] One possible implementation involves the adjustment module dynamically adjusting the weighting coefficients of risk dimension indicators based on the vehicle's risk dimension indicators and historical risk dimension indicators. Specifically, this involves: determining the indicator error corresponding to the risk dimension indicator based on the risk dimension indicator and its corresponding safety threshold; determining a correction value for the indicator error based on the indicator error and the historical indicator error corresponding to the vehicle's historical risk dimension indicators; determining a baseline value for the risk dimension indicator based on all risk dimension indicators within a preset time window; and correcting the baseline value using the correction value to obtain the dynamically adjusted weighting coefficients.

[0024] In one possible implementation, the vehicle real-time hazard assessment device provided in this application embodiment is further used to: normalize all dynamically adjusted weight coefficients to obtain the dynamically adjusted weight coefficients corresponding to each risk dimension index in the risk dimension index.

[0025] In one possible implementation, the vehicle real-time hazard assessment device provided in this application embodiment is further used to: determine whether the vehicle poses a driving hazard based on a real-time hazard index; and, if a driving hazard is determined, determine a vehicle control strategy based on the real-time hazard index to control the vehicle.

[0026] One possible implementation involves the module determining the vehicle's control strategy based on real-time hazard indicators, specifically by: maintaining the vehicle's autonomous driving state and recording data when the real-time hazard indicators are determined to be in a low-risk range; and / or generating a takeover warning and pre-tightening the seatbelts when the real-time hazard indicators are determined to be in a medium-risk range to remind the driver to take over the vehicle; and / or triggering the vehicle's emergency braking function or controlling the vehicle to perform a pullover operation when the determined real-time hazard indicators are determined to be in a high-risk range, and uploading accident data.

[0027] Thirdly, embodiments of this application provide a vehicle real-time hazard assessment device, which has the function of implementing the vehicle real-time hazard assessment method of the first aspect or any possible implementation thereof. This function can be implemented by hardware or by hardware executing corresponding software. The hardware or software includes one or more modules corresponding to the above-described function.

[0028] Fourthly, embodiments of this application provide a computer-readable storage medium storing instructions that, when executed on a computer, enable the computer to perform the vehicle real-time hazard assessment method described in the first aspect or any possible implementation thereof.

[0029] Fifthly, embodiments of this application provide a computer program product containing instructions that, when run on a computer, enable the computer to execute the vehicle real-time hazard assessment method described in the first aspect or any possible implementation thereof.

[0030] The technical effects of any of the design methods in aspects two through five can be found in aspect one or in different possible implementations of aspect one, and will not be repeated here. Attached Figure Description

[0031] To more clearly illustrate the technical solutions in the specific embodiments of this application or the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0032] Figure 1 A system structure diagram of a real-time vehicle hazard assessment system provided in this application embodiment; Figure 2 A flowchart of a method for real-time vehicle hazard assessment provided in this application embodiment; Figure 3 A schematic diagram of a vehicle real-time hazard assessment device provided in this application embodiment; Figure 4 Another system architecture diagram of a vehicle real-time hazard assessment system provided in this application embodiment. Detailed Implementation

[0033] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. The components of the embodiments of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0034] Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely to illustrate selected embodiments of the application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0035] The construction and calculation of performance evaluation metrics for autonomous vehicles is one of the core challenges in current technological development and commercialization. Related technologies have significant shortcomings in adapting to multi-dimensional needs and complex scenarios, mainly in the following aspects: First, regarding technological evolution and the need for anthropomorphism, as autonomous driving technology evolves from basic control functions to higher-level decision-making intelligence, the evaluation objective has shifted from simply "safety and stability" to a comprehensive evaluation of "human-like or even superhuman driving behavior." However, existing evaluation indicators are often too scenario-dependent, or micro-indicators and macro-statistical indicators are disconnected and difficult to unify, which seriously restricts the continuous optimization and iteration of the technology.

[0036] Secondly, regarding scenario complexity and evaluation standardization, autonomous driving systems need to cope with extremely complex and high-dimensional road operating environments. However, traditional testing methods, such as safety assessment methods based on preset collision scenarios, cannot comprehensively cover multiple performance dimensions such as comfort, traffic efficiency, and altruism. For example, while emergency braking can effectively avoid collisions, excessive braking intensity may significantly affect passenger comfort. This necessitates the coordinated evaluation of multiple indicators to balance different performance requirements. Furthermore, the varying complexity of different test scenarios leads to significant fluctuations in evaluation results, and existing technologies lack effective scenario complexity modeling mechanisms to achieve adaptive scaling of evaluation results.

[0037] Based on this, embodiments of this application provide a method and apparatus for real-time vehicle hazard assessment. The method includes acquiring vehicle operating data and environmental data of the vehicle's driving environment. Based on the operating data and environmental data, risk dimension indicators for the vehicle in the current driving environment are determined. The weighting coefficients of the risk dimension indicators are dynamically adjusted based on the vehicle's risk dimension indicators and historical risk dimension indicators. The real-time hazard index of the vehicle is calculated based on the dynamically adjusted weighting coefficients and the risk dimension indicators.

[0038] This application provides a method for real-time vehicle hazard assessment. By acquiring vehicle operational data and environmental data of the vehicle's driving environment, a risk dimension index for the vehicle in the current driving environment is determined based on this data. The weight of this risk dimension index is then dynamically adjusted using the vehicle's risk dimension index and its historical risk dimension index. This allows the weight coefficient to be dynamically adjusted according to the actual driving scenario, improving the adaptability and accuracy of the real-time vehicle hazard assessment under different road types, traffic conditions, and weather conditions. Finally, based on the dynamically adjusted weight coefficients and risk dimension index, a real-time hazard index for the vehicle in the current environment is calculated to assist in the autonomous driving control of the vehicle, providing a more reliable control basis for the decision-making and control of autonomous driving.

[0039] The methods provided in the embodiments of this application will now be described in conjunction with the specific accompanying drawings.

[0040] On the one hand, embodiments of this application provide a real-time vehicle hazard assessment system. For example... Figure 1 As shown, the vehicle real-time hazard assessment system 100 may include an on-board sensor 101, an on-board computing platform 102, a vehicle control domain controller 103, an actuator system 104, and an on-board data recorder 105.

[0041] The vehicle-mounted sensor 101 may include millimeter-wave radar arrays arranged on the front and rear bumpers and four corners of the vehicle, a lidar mounted on the roof, a stereo vision camera positioned above the windshield, and an inertial measurement unit and GPS / IMU integrated navigation system located at the vehicle's center of gravity. The vehicle-mounted sensor 101 is used to continuously acquire the vehicle's own operating status parameters such as speed, acceleration, and yaw rate during vehicle operation, while simultaneously detecting the position, speed, and trajectory of surrounding traffic participants, and providing the system with accurate vehicle positioning information.

[0042] The vehicle-mounted computing platform 102 can be a multi-core processor compliant with automotive-grade standards and equipped with a GPU acceleration module. This platform can process the raw data input from sensors in real time using the vehicle real-time hazard assessment method provided in this application embodiment. Specifically, it calculates instantaneous safety risk indicators, traffic efficiency indicators, and ride comfort indicators, and dynamically adjusts the weighting coefficients of each dimension indicator based on the current time and historical risk data. Finally, it completes the weighted fusion of multi-dimensional indicators and outputs the vehicle's real-time hazard assessment value.

[0043] The vehicle control domain controller 103 can establish a communication connection with the computing platform via CAN bus and in-vehicle Ethernet. This controller pre-stores threshold parameters corresponding to different risk levels. Upon receiving real-time hazard indicators from the computing platform, it immediately determines the risk level and generates corresponding control commands based on the determination result. In low-risk situations, only data logging commands are generated; in medium-risk situations, warning and seatbelt pretensioning commands are generated; and in high-risk situations, active control commands such as emergency braking or pulling over are generated.

[0044] The actuator system 104 may consist of multiple vehicle-level actuators, including an electronic stability control system responsible for maintaining stable vehicle operation, an electric braking system for precise braking force distribution, an electric power steering system for performing steering operations, and a seatbelt pretensioner for occupant restraint. After receiving instructions from the vehicle control domain controller 103, the actuator system 104 translates them into specific vehicle actions to achieve proactive intervention in risk mitigation.

[0045] The vehicle data recorder 105 uses automotive-grade storage devices to continuously record various parameters and risk indicators during system operation. The device features event-triggered storage; when the system detects a high-risk situation, it automatically saves complete data records for a period before and after the incident and uploads key data to a cloud service platform via the vehicle communication module. It also provides historical data support for long-term optimization of weighting coefficients.

[0046] It should be noted that the above Figure 1 The illustrated real-time vehicle hazard assessment system 100 is merely an example of the application scenario of this application solution and is not intended to limit the application scenario of this application solution.

[0047] On the one hand, embodiments of this application provide a method for real-time vehicle hazard assessment, which can be performed by... Figure 1 The vehicle real-time hazard assessment system 100 shown is running. (As shown) Figure 2 As shown, the method may include the following steps.

[0048] S201, acquire vehicle operating data and environmental data of the vehicle's driving environment.

[0049] The vehicle's operational data may include one or more of the following: longitudinal speed, lateral speed, yaw rate, steering wheel angle, throttle opening, braking pressure, wheel speeds, three-axis acceleration, three-axis angular velocity, positioning information, and heading angle. The environmental data of the vehicle's driving environment may include one or more of the following: relative distances, relative speeds, and azimuths of surrounding traffic participants; point cloud data of the surrounding environment; road image data ahead; traffic flow information broadcast from roadside units and other vehicles; meteorological parameters such as ambient temperature, humidity, and rainfall; road speed limits; road curvature data; and traffic congestion status.

[0050] One possible implementation involves collecting real-time data from the vehicle's chassis system via the onboard CAN bus, such as longitudinal speed, lateral speed, yaw rate, steering wheel angle, throttle opening, brake pressure, and wheel speeds. An inertial measurement unit (IMU) is used to collect the vehicle's three-axis acceleration and angular velocity. A GPS / IMU integrated navigation system is then used to obtain the vehicle's precise positioning information and heading angle.

[0051] One possible implementation involves using millimeter-wave radar arrays deployed around the vehicle to detect the relative distance, relative speed, and azimuth of surrounding road users. Point cloud data of the surrounding environment is acquired using lidar for constructing a high-precision local map and identifying obstacle contours. Images of the road ahead are captured by stereo vision cameras, and lane line detection, traffic sign recognition, and traffic light status identification are performed using deep learning algorithms. Traffic flow information broadcast from roadside units and other vehicles is received via a V2X communication module.

[0052] One possible approach is to acquire current environmental meteorological parameters, such as ambient temperature, humidity, and rainfall, through vehicle-mounted meteorological sensors, and combine this with visual information collected by cameras to make a comprehensive judgment on the current weather conditions.

[0053] One possible approach is to obtain road environment parameters such as road speed limit information, road curvature data, and traffic congestion status from the vehicle navigation system.

[0054] S202 determines the risk dimension indicators of the vehicle in the current driving environment based on operational and environmental data.

[0055] Among them, risk dimension indicators can include instantaneous safety risk indicators, traffic efficiency indicators, and ride comfort indicators.

[0056] One possible implementation, where the risk dimension indicator is an instantaneous safety risk indicator, involves determining the vehicle's risk dimension indicator in the current driving environment based on operational and environmental data. This includes: for any target traffic participant among all traffic participants surrounding the vehicle, determining the repulsive potential field between the vehicle and the target traffic participant based on the vehicle's current position in the operational data and the target traffic participant's current position in the environmental data. The repulsive potential field is then corrected using spatial and temporal risk sensitivity to obtain the target instantaneous safety risk indicator between the vehicle and the target traffic participant. The maximum value among the target instantaneous safety risk indicators of all target traffic participants surrounding the vehicle is determined as the instantaneous safety risk indicator.

[0057] Specifically, for each traffic participant detected around the vehicle, a repulsive potential field model is established between the vehicle's current position and the traffic participant's current position. This repulsive potential field is effective within a set safe distance, and its strength is inversely proportional to the relative distance. A spatial risk sensitivity parameter is introduced to spatially correct the repulsive potential field. This parameter uses an exponential decay function to characterize the influence of distance on risk perception; a larger sensitivity coefficient indicates lower sensitivity to long-distance risks. Simultaneously, a temporal risk sensitivity parameter is introduced for temporal correction. This parameter also uses an exponential decay function to characterize the time decay characteristics of risks over future time periods; a larger sensitivity coefficient indicates lower sensitivity to long-term risks. The repulsive potential field value after both spatial and temporal corrections is used as the target instantaneous safety risk indicator corresponding to that traffic participant. All detected traffic participants are traversed, and the largest target instantaneous safety risk indicator value is taken as the final instantaneous safety risk indicator output.

[0058] For example, starting from the definition and essence of driving risk, the definition of driving risk is further refined to "the mapping of collision damage to the present state through uncertainty modeling". By introducing the artificial potential field method and combining it with spatial and temporal reduction functions for modeling, the current collision risk is quantified. Its expression is as follows:

[0059] in, It serves as an indicator of instantaneous safety risks.

[0060] Gnt is the damage estimate for a predicted collision with the nth traffic participant at time t. Gnt can be calculated by introducing a repulsive potential field function, which is effective within a safe distance ρ0 from the obstacle (surrounding road traffic participants). The expression for Gnt is as follows:

[0061] Where krep is the repulsive force gain coefficient, and d(q,qobs) is the distance from the current position to the obstacle; fs(dnt) is a spatial reduction function that reflects the impact of distance on risk. fs(dnt) can be determined by the following formula.

[0062]

[0063] Where d is the distance between the vehicle and the object's outline. B is the spatial risk sensitivity. The larger B is, the less sensitive it is to long-distance risks.

[0064] ft(T t) is a time reduction function that predicts the time decay of risk within the time domain T. t) can be determined by the following formula.

[0065]

[0066] Here, A represents time risk sensitivity. The larger A is, the less sensitive it is to long-term risks.

[0067] Another possible implementation, where the risk dimension indicator is a traffic efficiency indicator, involves determining the vehicle's risk dimension indicator in the current driving environment based on operational and environmental data. This includes: determining the vehicle's average speed based on the vehicle's speed at each moment within a preset time window from the operational data; obtaining the average speed of the traffic flow within the preset time window from the environmental data; and determining the traffic efficiency indicator based on the vehicle's average speed, the average traffic flow speed, and the weather impact coefficient corresponding to the vehicle's environment from the environmental data.

[0068] Specifically, from the perspective of the overall transportation system, individual optimality does not necessarily represent overall optimality. For example, frequent lane changes may improve the efficiency of a single vehicle, but they also frequently disrupt traffic flow, increasing road traffic risks and affecting overall road traffic efficiency. Furthermore, vehicle speed is affected by traffic conditions; for instance, speeds will decrease during congestion. This is addressed by introducing the traffic flow deviation index IDE(t) (the ratio of the vehicle's speed to the average speed of surrounding traffic) and incorporating environmental and weather conditions for auxiliary correction.

[0069] When calculating the average traffic speed in the surrounding area, vehicles within a certain distance in front of and behind the vehicle, or within the sensor's perception range, should be considered. These vehicles should have similar travel routes to eliminate the influence of oncoming vehicles and vehicles in different turning lanes at intersections. If there are no other vehicles in the vicinity, the current road speed limit can be used as the average traffic speed. Using the ratio to the average speed of the surrounding traffic flow as an evaluation metric has the advantage of being suitable for reasonable assessment in low-speed scenarios such as congestion and traffic light stops. It considers the impact of the surrounding traffic environment with minimal computational complexity, avoiding misjudgments that may result from solely pursuing speed.

[0070] For example, traffic efficiency indicators can be determined using the following formula.

[0071]

[0072] in, Here, vi is the traffic efficiency index, vi is the i-th sampled speed of the vehicle within the time window t, vtraffic is the average speed of the surrounding traffic flow, and α(W) is the weather influence coefficient, defined in segments according to meteorological levels.

[0073] As shown in Table 1, the weather impact coefficient α(W) can be segmented according to real-time meteorological conditions.

[0074] Table 1

[0075] Another possible implementation, where the risk dimension indicator is the ride comfort indicator, involves determining the vehicle's risk dimension indicator in the current driving environment based on operational and environmental data. This includes: determining the instantaneous acceleration of the vehicle along multiple preset axes in the vehicle's coordinate system based on the vehicle's acceleration in the operational data; weighting and filtering the target instantaneous acceleration for any one of the preset axes to obtain the weighted and filtered target instantaneous acceleration; calculating the root mean square value of the weighted and filtered target instantaneous acceleration; and weighted and fusing the root mean square values ​​of the target instantaneous accelerations along all preset axes to obtain the ride comfort indicator.

[0076] For example, when calculating the ride comfort index, based on the ISO 2631-1:1997 standard, the root mean square (RMS) method is used to combine and calculate the three-axis acceleration in the vehicle coordinate system to measure driving comfort. Before calculating the RMS, frequency weighting is used to filter vibrations of different frequencies, making the calculation results closer to human perception. The weighted acceleration is used to calculate the RMS of each axis's acceleration (awx, awy, awz) within time period T; then, the overall weighted root mean square (OVTV) is further calculated from the three-axis acceleration RMS, and driving comfort is judged based on this value. The specific implementation steps are as follows: First, instantaneous accelerations along three axes are obtained in the vehicle coordinate system: X-axis (chest and back direction), Y-axis (shoulder and arm direction), and Z-axis (vertical direction). Frequency-weighted filtering is then applied to the instantaneous accelerations in each direction. Specifically, for Z-axis acceleration, the Wk weighting function is used, with a sensitive frequency range of 4-12.5Hz, which mainly reflects the response of the human internal organs and spine to vertical vibration.

[0077] For X-axis and Y-axis acceleration, the Wd weighting function is used, with a sensitive frequency range of 0.5-2Hz, which is more in line with the human body's sensitivity to horizontal vibrations.

[0078] Next, calculate the weighted root mean square (RMS) acceleration values ​​for each axis. The calculation formula is as follows:

[0079] Where a(t) is the instantaneous acceleration, w(f) is the frequency weighting function (Wk or Wd) in the corresponding direction, and T is the measurement duration.

[0080] Finally, in a seated position, multiply the weighted accelerations along the X and Y axes by a direction factor of 1.4, and the weighted acceleration along the Z axis by a direction factor of 1.0. Combine the weighted accelerations in the three directions using the following formula:

[0081] in, For ride comfort indicators, The weighted root mean square value of acceleration along the X-axis. The weighted root mean square acceleration along the Y-axis, This is the weighted root mean square value of the acceleration along the Z-axis.

[0082] S203 dynamically adjusts the weighting coefficients of the risk dimension indicators based on the vehicle's risk dimension indicators and the vehicle's historical risk dimension indicators.

[0083] Specifically, based on the risk dimension indicators and their corresponding safety thresholds, the indicator errors for each risk dimension indicator are determined. Based on these errors and the historical indicator errors corresponding to the vehicle's historical risk dimension indicators, correction values ​​for the indicator errors are determined. Based on all risk dimension indicators within a preset time window, baseline values ​​for the risk dimension indicators are determined. The baseline values ​​are then adjusted using the correction values ​​to obtain dynamically adjusted weighting coefficients.

[0084] One possible implementation involves dynamically adjusting the weighting coefficients through a dual-loop adaptive mechanism, which includes a real-time rapid response in the inner loop and a long-term stable calibration in the outer loop. First, the index error is calculated based on the deviation between the real-time values ​​of each risk dimension indicator and its corresponding safety threshold. Specifically, the instantaneous safety risk error is defined as the difference between the current safety risk indicator and the safety threshold, while the traffic efficiency error and ride comfort error are calculated from the deviations of their respective indicator values ​​from their corresponding thresholds.

[0085] During the inner-loop adjustment process, an incremental PID control algorithm is used to correct the weight coefficients in real time. Taking the instantaneous safety risk weight w1 as an example, its weight adjustment is calculated through a combination of proportional, integral, and derivative terms: the proportional term reflects the direct impact of the current error, the integral term accumulates historical errors to eliminate steady-state errors, and the derivative term predicts the error change trend to suppress overshoot. The weight coefficients are then updated based on the adjustment amount, ensuring that autonomous vehicles can respond quickly to sudden risks.

[0086] During the outer-loop calibration process, historical average values ​​of each risk dimension indicator are calculated using a sliding window statistical method. Based on this historical data, an objective function is constructed to minimize the deviation between the long-term average hazard and the system's expected target value, while a regularization term is introduced to prevent weight overfitting. Gradient descent is used to optimize the baseline weight values, ensuring the vehicle maintains global stability under different driving scenarios.

[0087] Furthermore, after obtaining the dynamically adjusted weight coefficients, all dynamically adjusted weight coefficients are normalized to obtain the dynamically adjusted weight coefficients corresponding to each risk dimension indicator in the risk dimension indicators.

[0088] One possible implementation involves normalizing the adjusted weights to ensure consistency among the weight coefficients, so that the sum of the three is always equal to 1. This mechanism guarantees both the real-time nature of weight adjustments and long-term operational stability, effectively improving the system's adaptability in different scenarios.

[0089] For example, The weighting coefficients corresponding to instantaneous safety risk indicators. These are the weighting coefficients corresponding to traffic efficiency indicators. The weighting coefficients corresponding to the ride comfort index.

[0090] Adaptive adjustment of weight parameters is key to improving model flexibility and scenario adaptability. This application employs a dual-loop adaptive parameter tuning method, which adjusts weights through coordinated inner and outer loop adjustments, balancing real-time performance and stability to achieve dynamic weight adjustment. The specific formulas and steps are as follows: (1) During the inner loop adjustment process, weights are rapidly adjusted based on real-time data to suppress sudden risks. The inner loop employs incremental PID control, adjusting weights according to instantaneous errors. For example: Definition of instantaneous safety risk index error:

[0091] in, This refers to the index error of the instantaneous safety risk indicator. As a safety threshold, The value can be customized. It serves as an indicator of instantaneous safety risks.

[0092] Definition of error for traffic efficiency index:

[0093] in, This represents the error in traffic efficiency indicators. As a safety threshold, The value can be customized. It is a traffic efficiency indicator.

[0094] Definition of error for ride comfort index:

[0095] in, This represents the error in the ride comfort index. As a safety threshold, The value can be customized. This is an indicator of passenger comfort.

[0096] Then, weight adjustment is performed using incremental PID:

[0097] Where Kp, Ki, and Kd are PID parameters, which can be calibrated experimentally, and Δt is the sampling period.

[0098] Finally, the weights are updated.

[0099] It should be noted that the above is only based on The adjustment process is shown in the example. Adjustment as well as The process and The process is the same, and will not be elaborated upon here. However, normalization processing is required: + + =1.

[0100] (2) During the outer loop calibration process, the benchmark weights are calibrated based on historical data to ensure global stability. The outer loop uses a sliding window to collect historical data and optimize the benchmark weight values.

[0101] First, determine the average value of instantaneous safety risk indicators:

[0102] in, This represents the average value of instantaneous safety risk indicators.

[0103] Then, determine the average traffic efficiency index:

[0104] in, This represents the average value of traffic efficiency indicators.

[0105] Furthermore, determine the average value of the ride comfort index:

[0106] in, This represents the average value of the ride comfort index.

[0107] Then, the outer loop objective function is constructed. First, determine how to minimize the long-term risk error: J =

[0108] Where J represents minimizing the long-term risk error. =w1* +w2* + w3* . This represents the expected long-term average level of risk for the autonomous driving system, and its value indicates the risk tolerance. λ is the regularization coefficient.

[0109] Then, the gradient descent method is used to update the baseline weight values.

[0110]

[0111] in, This is the baseline value for the weighting.

[0112] It should be noted that the above is only based on The adjustment process is shown in the example. Adjustment as well as The process and The process is the same, and will not be described in detail here.

[0113] Finally, the inner ring is rapidly adjusted based on the benchmark value provided by the outer ring to determine the weighting coefficients.

[0114] , , .

[0115] S204 calculates the real-time hazard index of a vehicle based on dynamically adjusted weighting coefficients and risk dimension indicators.

[0116] For example, after obtaining the dynamically adjusted weighting coefficients , as well as Then, a weighted fusion method is used to calculate the real-time hazard index for autonomous driving. This real-time hazard index integrates three core dimensions: instantaneous safety risk index, traffic efficiency index, and ride comfort index. Its mathematical expression is:

[0117] in, , , These are the weighting coefficients; As an instantaneous safety risk indicator at the current moment, As a traffic efficiency indicator. This is an indicator of passenger comfort.

[0118] Furthermore, based on real-time hazard indicators, it is determined whether the vehicle poses a driving hazard. If a driving hazard is determined, a vehicle control strategy is determined based on the real-time hazard indicators to control the vehicle.

[0119] For example, the vehicle control strategy can be determined by the following Table 2.

[0120] Table 2

[0121] One possible approach is to maintain the vehicle's autonomous driving state and record data when the real-time hazard index is determined to be in a low-risk range.

[0122] For example, when ADRI(t)∈[0, 0.3], it is determined to be a low-risk level. In this state, the vehicle maintains its current autonomous driving mode without active intervention, but continuously records vehicle operation data, environmental data, and various risk dimension indicators for subsequent system optimization and analysis.

[0123] Another possible approach is to generate a takeover warning and pre-tighten the seatbelts when the real-time hazard index is determined to be in the medium-risk range, in order to remind the driver to take over the vehicle.

[0124] For example, when ADRI(t) ∈ (0.3, 0.6], it is determined to be at a medium risk level. The warning mechanism is immediately activated, and a takeover warning is issued to the driver through a multimodal human-machine interface such as vision, hearing, and touch. At the same time, the seat belts are pre-tightened to prepare for potential collision risks, and the control authority of the autonomous driving system is gradually reduced to create transitional conditions for driver takeover.

[0125] Another possible implementation is to trigger the vehicle's emergency braking function or control the vehicle to pull over to the side of the road when the determined real-time hazard index is in the high-risk range, and upload the accident data.

[0126] For example, when ADRI(t) ∈ (0.6, 1], it is determined to be a high-risk level. The highest level of safety response is immediately triggered: first, the emergency braking system is activated, and maximum deceleration braking is performed on the premise of ensuring safety behind; if environmental conditions permit, the vehicle is controlled to perform a pullover operation. At the same time, the accident black box function is automatically activated, and sensor data, system status, and decision records within the key time windows before and after the incident are packaged and uploaded to the cloud service platform, preserving a complete chain of evidence for subsequent fault analysis and liability determination.

[0127] The above primarily describes the solutions provided in this application from the perspective of the device's working principle. It is understood that, to achieve the aforementioned functions, the vehicle real-time hazard assessment device includes corresponding hardware structures and / or software modules for executing each function. Those skilled in the art should readily recognize that, based on the algorithm steps of the examples described in the embodiments disclosed herein, this application can be implemented in hardware or a combination of hardware and computer software. Whether a function is executed in hardware or by computer software driving hardware depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0128] This application embodiment can divide the vehicle real-time hazard assessment device into functional modules according to the above method example. For example, each function can be divided into a separate functional module, or two or more functions can be integrated into one processing module. The integrated module can be implemented in hardware or as a software functional module.

[0129] It should be noted that the module division in this embodiment is illustrative and represents only one logical functional division; in actual implementation, other division methods may be used. When dividing functional modules according to their respective functions, Figure 3 A schematic diagram of a possible composition of the vehicle real-time hazard assessment device involved in the above and embodiment examples is shown. Figure 3 As shown, the vehicle real-time hazard assessment device 300 may include: an acquisition module 301, a determination module 302, an adjustment module 303, and a calculation module 304.

[0130] The acquisition module 301 is used to support the execution of the vehicle real-time hazard assessment device 300. Figure 2 S201 in the illustrated method for real-time vehicle hazard assessment.

[0131] The determination module 302 is used to support the execution of the vehicle real-time hazard assessment device 300. Figure 2 S202 in the illustrated method for real-time vehicle hazard assessment.

[0132] Adjustment module 303 is used to support the execution of the vehicle real-time hazard assessment device 300. Figure 2 S203 in the illustrated method for real-time vehicle hazard assessment.

[0133] Calculation module 304 is used to support the execution of the vehicle real-time hazard assessment device 300. Figure 2 S204 in the illustrated method for real-time vehicle hazard assessment.

[0134] One possible implementation involves using environmental data including at least one traffic participant surrounding the vehicle. When the risk dimension indicator is an instantaneous safety risk indicator, the determination module, when determining the vehicle's risk dimension indicator in the current driving environment based on operational data and environmental data, specifically performs the following: For any target traffic participant among all traffic participants surrounding the vehicle, based on the vehicle's current position in the operational data and the target traffic participant's current position in the environmental data, determine the repulsive potential field between the vehicle and the target traffic participant. The repulsive potential field is then corrected using spatial risk sensitivity and temporal risk sensitivity to obtain the target instantaneous safety risk indicator between the vehicle and the target traffic participant. The maximum value among the target instantaneous safety risk indicators of all target traffic participants surrounding the vehicle is determined as the instantaneous safety risk indicator.

[0135] One possible implementation, where the risk dimension indicator is a traffic efficiency indicator, involves the following steps when the module determines the risk dimension indicator of a vehicle in the current driving environment based on operational and environmental data: First, it determines the vehicle's average speed based on the vehicle's speed at each moment within a preset time window from the operational data. Second, it obtains the average speed of the traffic flow within the preset time window from the environmental data. Finally, it determines the traffic efficiency indicator based on the vehicle's average speed, the average traffic flow speed, and the weather influence coefficient corresponding to the vehicle's environment from the environmental data.

[0136] One possible implementation, where the risk dimension indicator is a ride comfort indicator, involves the following steps when the determination module, based on operational and environmental data, determines the vehicle's risk dimension indicator in the current driving environment: First, based on the vehicle's acceleration in the operational data, determine the instantaneous acceleration of the vehicle along multiple preset axes in the vehicle's coordinate system. Then, for any one of these preset axes, perform a weighted filtering on the target instantaneous acceleration to obtain a weighted filtered target instantaneous acceleration. Next, calculate the root mean square (RMS) value of the weighted filtered target instantaneous acceleration. Finally, weightedly fuse the RMS values ​​of the target instantaneous accelerations along all preset axes to obtain the ride comfort indicator.

[0137] One possible implementation involves the adjustment module dynamically adjusting the weighting coefficients of risk dimension indicators based on the vehicle's risk dimension indicators and historical risk dimension indicators. Specifically, this involves: determining the indicator error corresponding to the risk dimension indicator based on the risk dimension indicator and its corresponding safety threshold; determining a correction value for the indicator error based on the indicator error and the historical indicator error corresponding to the vehicle's historical risk dimension indicators; determining a baseline value for the risk dimension indicator based on all risk dimension indicators within a preset time window; and correcting the baseline value using the correction value to obtain the dynamically adjusted weighting coefficients.

[0138] In one possible implementation, the vehicle real-time hazard assessment device provided in this application embodiment is further used to: normalize all dynamically adjusted weight coefficients to obtain the dynamically adjusted weight coefficients corresponding to each risk dimension index in the risk dimension index.

[0139] In one possible implementation, the vehicle real-time hazard assessment device provided in this application embodiment is further used to: determine whether the vehicle poses a driving hazard based on a real-time hazard index; and, if a driving hazard is determined, determine a vehicle control strategy based on the real-time hazard index to control the vehicle.

[0140] One possible implementation involves the module determining the vehicle's control strategy based on real-time hazard indicators, specifically by: maintaining the vehicle's autonomous driving state and recording data when the real-time hazard indicators are determined to be in a low-risk range; and / or generating a takeover warning and pre-tightening the seatbelts when the real-time hazard indicators are determined to be in a medium-risk range to remind the driver to take over the vehicle; and / or triggering the vehicle's emergency braking function or controlling the vehicle to perform a pullover operation when the determined real-time hazard indicators are determined to be in a high-risk range, and uploading accident data.

[0141] It should be noted that all relevant content of each step involved in the above method embodiments can be referenced from the functional description of the corresponding functional module, and will not be repeated here.

[0142] The vehicle real-time hazard assessment device 300 provided in this application embodiment is used to perform the above-mentioned... Figure 2 The vehicle real-time hazard assessment method shown can therefore achieve the same effect as the vehicle real-time hazard assessment method described above.

[0143] This application also provides a vehicle real-time hazard assessment device, which can perform the vehicle real-time hazard assessment method and related steps described in the above method embodiments.

[0144] This application also provides a computer-readable storage medium storing instructions that, when executed, perform the vehicle real-time hazard assessment method and related steps in the above method embodiments.

[0145] This application also provides a computer program product that, when run on a computer, causes the computer to execute the vehicle real-time hazard assessment method and related steps described in the above method embodiments.

[0146] In some embodiments, the methods shown in this application can be implemented as computer program instructions encoded in a machine-readable format on a computer-readable storage medium or on other non-transitory media or articles of art.

[0147] This application also provides a vehicle real-time hazard assessment system 100, such as... Figure 4 As shown, the vehicle real-time hazard assessment system 100 includes at least one processor 401 and at least one interface circuit 402.

[0148] As an example, when the vehicle real-time hazard assessment system 100 includes a processor and an interface circuit, then the processor can be... Figure 4 The processor 401 shown in the solid box (or the processor 401 shown in the dashed box) can be an interface circuit. Figure 4 The interface circuit 402 is shown in the solid box (or the interface circuit 402 shown in the dashed box). When the vehicle real-time hazard assessment system 100 includes two processors and two interface circuits, then the two processors include... Figure 4 The processor 401 shown in the solid box and the processor 401 shown in the dashed box, these two interface circuits include Figure 4 Interface circuit 402 is shown in both solid and dashed boxes. No limitations are imposed on this.

[0149] Processor 401 and interface circuit 402 can be interconnected via a line. For example, interface circuit 402 can be used to receive signals. Alternatively, interface circuit 402 can be used to send signals to other devices (e.g., processor 401). For instance, interface circuit 402 can read computer instructions stored in memory and send those instructions to processor 401. Processor 401 executes the instructions and, in conjunction with input / output devices, implements the various steps in the above embodiments, such as implementing... Figure 2 The methods illustrated are the steps performed in the embodiments shown. Of course, this real-time vehicle hazard assessment system may also include other discrete components, and this application does not specifically limit this.

[0150] Through the above description of the embodiments, those skilled in the art can clearly understand that, for the sake of convenience and brevity, only the division of the above functional modules is used as an example. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above.

[0151] In the several embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of modules or units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another device, or some features may be ignored or not executed. Furthermore, the mutual coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0152] The units described as separate components may or may not be physically separate. A component shown as a unit can be one or more physical units; that is, it can be located in one place or distributed in multiple different locations. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0153] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0154] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a readable storage medium. Based on this understanding, the technical solution of the embodiments of this application, or the part that contributes to it, or all or part of the technical solution, can be embodied in the form of a software product. This software product is stored in a storage medium and includes several instructions to cause a device (which may be a microcontroller, chip, etc.) or processor to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0155] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for real-time vehicle hazard assessment, characterized in that, The method includes: Acquire vehicle operating data and environmental data of the vehicle's driving environment; Based on the operational data and the environmental data, determine the risk dimension indicators of the vehicle in the current driving environment; The weighting coefficients of the risk dimension indicators are dynamically adjusted based on the risk dimension indicators of the vehicle and the historical risk dimension indicators of the vehicle. The real-time hazard index of the vehicle is obtained by calculating based on the dynamically adjusted weighting coefficients and the risk dimension indicators.

2. The method according to claim 1, characterized in that, The environmental data includes at least one traffic participant around the vehicle; when the risk dimension indicator is an instantaneous safety risk indicator, determining the risk dimension indicator of the vehicle in the current driving environment based on the operational data and the environmental data includes: For any target traffic participant among all traffic participants around the vehicle, the repulsive potential field between the vehicle and the target traffic participant is determined based on the current position of the vehicle in the operational data and the current position of the target traffic participant in the environmental data. By modifying the repulsive potential field using spatial risk sensitivity and temporal risk sensitivity, a target instantaneous safety risk index between the vehicle and the target traffic participant is obtained. The maximum value among the target instantaneous safety risk indicators of all target traffic participants around the vehicle is determined as the instantaneous safety risk indicator.

3. The method according to claim 1, characterized in that, When the risk dimension indicator is a traffic efficiency indicator, determining the risk dimension indicator of the vehicle in the current driving environment based on the operational data and the environmental data includes: The average speed of the vehicle is determined based on the vehicle's speed at each moment within a preset time window in the operational data. Obtain the average speed of the traffic flow where the vehicle is located within the preset time window from the environmental data; The traffic efficiency index is determined based on the average vehicle speed, the average traffic flow speed, and the weather influence coefficient corresponding to the environment in which the vehicle is located in the environmental data.

4. The method according to claim 1, characterized in that, When the risk dimension indicator is a ride comfort indicator, determining the risk dimension indicator of the vehicle in the current driving environment based on the operating data and the environmental data includes: Based on the vehicle's acceleration in the operational data, determine the instantaneous acceleration of the vehicle along multiple preset axes in the vehicle body coordinate system; For the instantaneous acceleration of a target along any one of a plurality of preset axes, the instantaneous acceleration of the target is weighted and filtered to obtain the weighted and filtered instantaneous acceleration of the target. Calculate the root mean square value of the instantaneous acceleration of the target after weighted filtering; The root mean square values ​​of the instantaneous accelerations of all preset axes are weighted and fused to obtain the ride comfort index.

5. The method according to claim 1, characterized in that, The step of dynamically adjusting the weighting coefficients of the risk dimension indicators based on the vehicle's risk dimension indicators and the vehicle's historical risk dimension indicators includes: Based on the risk dimension indicators and the corresponding safety thresholds, the indicator errors corresponding to the risk dimension indicators are determined. Based on the index error and the historical index error corresponding to the historical risk dimension index of the vehicle, the correction value of the index error is determined; Based on all risk dimension indicators within a preset time window, determine the baseline value of the risk dimension indicator; The benchmark value is corrected using the correction value to obtain the dynamically adjusted weighting coefficient.

6. The method according to claim 5, characterized in that, The method further includes: All dynamically adjusted weight coefficients are normalized to obtain the dynamically adjusted weight coefficients for each risk dimension indicator.

7. The method according to claim 1, characterized in that, The method further includes: Based on the real-time hazard index, determine whether the vehicle poses a driving hazard; If it is determined that the vehicle poses a driving hazard, a control strategy for the vehicle is determined based on the real-time hazard index in order to control the vehicle.

8. The method according to claim 7, characterized in that, The step of determining the vehicle control strategy based on the real-time hazard index includes: If the real-time hazard index is determined to be in a low-risk range, the vehicle's autonomous driving state is maintained and data is recorded; And / or, if the real-time hazard index is determined to be in the medium-risk range, a takeover warning is generated and the seat belts are pre-tightened to remind the driver to take over the vehicle; And / or, if the determined real-time hazard index is in the high-risk range, trigger the vehicle's emergency braking function or control the vehicle to perform a pull-over operation, and upload the accident data.

9. A vehicle real-time hazard assessment device, characterized in that, The device includes: The acquisition module is used to acquire the vehicle's operating data and the environmental data of the vehicle's driving environment. The determination module is used to determine the risk dimension indicators of the vehicle in the current driving environment based on the operating data and the environmental data; The adjustment module is used to dynamically adjust the weight coefficients of the risk dimension indicators based on the risk dimension indicators of the vehicle and the historical risk dimension indicators of the vehicle. The calculation module is used to calculate the real-time hazard index of the vehicle based on the dynamically adjusted weight coefficients and the risk dimension indicators.

10. A vehicle real-time hazard assessment device, characterized in that, The vehicle real-time hazard assessment device includes a processor and a memory. The memory stores machine-executable instructions that can be executed by the processor. The processor executes the machine-executable instructions to implement the vehicle real-time hazard assessment method according to any one of claims 1 to 8.

Images