Driver state evaluation method and system, and car-road cooperation system

Abstract

The application discloses a driver state evaluation method and system and a vehicle-road cooperative system. The method comprises the following steps: acquiring vehicle passing records collected by a plurality of vehicle-road cooperative communication devices arranged along a road; forming a space-time trajectory or a space-time trajectory segment of a target vehicle according to the vehicle passing records of the target vehicle, and determining driving characteristics of the target vehicle in the space-time trajectory or the space-time trajectory segment; determining vehicles satisfying preset peer conditions according to the vehicle passing records to construct a peer reference vehicle group; determining baseline driving characteristics of the peer reference vehicle group; determining difference characteristics of the target vehicle relative to the peer reference vehicle group according to the driving characteristics of the target vehicle and the baseline driving characteristics of the peer reference vehicle group; and inputting at least the difference characteristics into a preset driver state evaluation model to output a driver state evaluation result corresponding to the target vehicle. The driver state evaluation method according to the embodiment of the application can improve the accuracy and reliability of driver state evaluation.

Description

Technical Field

[0001] This invention relates to the field of road traffic, and in particular to a driver status assessment method and system, and a vehicle-road cooperative system. Background Technology

[0002] In the field of road transportation, the driver's condition has a significant impact on driving safety. Fatigue driving and other poor driving conditions are major contributing factors to traffic accidents. Therefore, solutions for monitoring driver condition have been proposed.

[0003] CN116494990A discloses an intelligent vehicle-mounted dashcam and alarm system for avoiding fatigue driving, which detects the driver's driving status through facial feature recognition and voice recognition.

[0004] CN110194174A discloses a fatigue driving monitoring system, which uses three infrared distance sensors to detect changes in the position of the head inside the vehicle, thereby determining whether the driver is fatigued.

[0005] However, the implementation of known solutions requires continuous monitoring of users' posture, voice, and even facial expressions, which inevitably poses a risk of privacy data leakage. The data security and compliance of these solutions are questionable.

[0006] Therefore, there is still room for improvement in the known methods for assessing driver condition.

[0007] The background description is provided for the purpose of understanding the relevant technologies in this field and is not intended as an admission of prior art. Summary of the Invention

[0008] The present invention aims to provide a driver state assessment method, a driver state assessment system, and a vehicle-road cooperative system, which can non-intrusively assess the driving state of the driver of a target vehicle and improve the accuracy and reliability of driver state assessment.

[0009] In a first aspect, embodiments of the present invention provide a driver state assessment method, comprising:

[0010] The vehicle passage records are collected by multiple vehicle-road cooperative communication devices set up along the road, wherein the vehicle passage records include at least vehicle identification, passage time and collection point location information;

[0011] Based on the vehicle passage records of the target vehicle, a spatiotemporal trajectory or spatiotemporal trajectory segment of the target vehicle is formed, and the driving characteristics of the target vehicle in the spatiotemporal trajectory or spatiotemporal trajectory segment are determined.

[0012] At least vehicles that meet the preset equivalence conditions are identified based on the vehicle passage records to construct an equivalence reference vehicle group;

[0013] Determine the baseline driving characteristics of the equivalent reference vehicle group;

[0014] Based on the driving characteristics of the target vehicle and the baseline driving characteristics of the equivalent reference vehicle group, the difference characteristics of the target vehicle relative to the equivalent reference vehicle group are determined.

[0015] The difference features are at least input into a preset driver state assessment model to output the driver state assessment result corresponding to the target vehicle.

[0016] In some embodiments, determining the driving characteristics of the target vehicle in the spatiotemporal trajectory or spatiotemporal trajectory segment includes:

[0017] The first driving characteristic of the target vehicle is calculated based on the distance between at least two collection points and the travel time of the target vehicle through the at least two collection points;

[0018] The determination of the baseline driving characteristics of the equivalent reference vehicle group includes:

[0019] The first reference driving characteristics of each peer reference vehicle are calculated based on the distance between the at least two collection points and the travel time of each peer reference vehicle between the at least two collection points.

[0020] The first baseline driving feature is obtained by statistically processing the first reference driving feature.

[0021] In some embodiments, determining the driving characteristics of the target vehicle in the spatiotemporal trajectory or spatiotemporal trajectory segment includes:

[0022] Calculate multiple first driving characteristics of the target vehicle based on the distance between at least three collection points and the travel time of the target vehicle through the at least three collection points;

[0023] Statistical processing is performed on the multiple first driving characteristics to obtain statistical driving characteristics;

[0024] Based on the plurality of first driving characteristics and the statistical driving characteristics, the second driving characteristics of the target vehicle are determined;

[0025] The determination of the baseline driving characteristics of the equivalent reference vehicle group includes:

[0026] Based on the distance between the at least three collection points and the travel time of each peer reference vehicle through the at least three collection points, multiple first reference driving characteristics of each peer reference vehicle are calculated.

[0027] The statistical reference driving characteristics of each peer reference vehicle are obtained by statistically processing the multiple first reference driving characteristics of each peer reference vehicle.

[0028] Based on the plurality of first reference driving characteristics and the statistical reference driving characteristics, the second reference driving characteristics of each equivalent reference vehicle are determined;

[0029] The second baseline driving characteristics are obtained by statistically processing the second reference driving characteristics.

[0030] In some embodiments, the first driving characteristic includes vehicle speed;

[0031] The difference features include vehicle speed deviation, which characterizes the relative deviation between the target vehicle's speed and the average speed of the equivalent reference vehicle group, and is determined by the following formula:

[0032]

[0033] in, For vehicle speed deviation, The average speed of the target vehicle between two adjacent data collection points. The baseline average speed of the equivalent reference vehicle group on the same road segment.

[0034] In some embodiments, the second driving characteristic includes driving smoothness, which characterizes the rate of change of vehicle speed across multiple consecutive road segments, and the driving smoothness of the target vehicle is determined by the following formula:

[0035]

[0036] in, For the driving smoothness of the target vehicle. Let be the average speed of the target vehicle in the j-th target road segment. For the target vehicle in Average vehicle speed on each road segment Number of road segments;

[0037] The driving smoothness of each equivalent reference vehicle is determined by the following formula:

[0038]

[0039] in, For the driving smoothness of each equivalent reference vehicle, Let be the interval average speed of each equivalent reference vehicle on the i-th target road segment. For the target vehicle in Average vehicle speed on each road segment Number of road segments;

[0040] The differential characteristics include a smoothness anomaly index, which is determined by the following formula:

[0041]

[0042] in, The smoothness anomaly index for the target vehicle. For the driving smoothness of the target vehicle. For the driving smoothness of each equivalent reference vehicle The average value.

[0043] In some embodiments, the driver state assessment method further includes:

[0044] Based on the degree of dispersion of the baseline driving characteristics of the equivalent reference vehicle group, traffic flow coordination characteristics are determined, wherein the traffic flow coordination characteristics are negatively correlated with the degree of dispersion.

[0045] The step of inputting the difference features into a preset driver state assessment model to output the driver state assessment result corresponding to the target vehicle includes: inputting the difference features and the traffic flow coordination feature into a preset driver state assessment model to output the driver state assessment result corresponding to the target vehicle.

[0046] In some embodiments, the driver state assessment method further includes:

[0047] In response to the driver status assessment result meeting the preset intervention conditions, an intervention operation is performed on the target vehicle and / or the driver of the target vehicle;

[0048] Obtain the vehicle passage records of the target vehicle at subsequent collection points, and determine the subsequent driver status assessment result of the target vehicle based on the subsequent vehicle passage records;

[0049] Based on the subsequent driver status assessment results, the effectiveness of the intervention is evaluated to obtain intervention events including effectiveness labels;

[0050] The driver status assessment model is updated based on the intervention event.

[0051] In some embodiments, determining vehicles that meet preset equivalence conditions to construct an equivalent reference vehicle group includes:

[0052] The vehicle reference vehicle group is constructed by filtering vehicles from the vehicle traffic records that are the same or similar to the target vehicle in terms of vehicle type, road segment, and direction of travel, and that pass through the same collection point within a preset time window.

[0053] In a second aspect, embodiments of the present invention provide a driver state assessment system, comprising:

[0054] The acquisition module is configured to acquire vehicle passage records collected by multiple vehicle-road cooperative communication devices set up along the road, wherein the passage records include at least vehicle identification, passage time and collection point location information;

[0055] The first determining module is configured to form the spatiotemporal trajectory or spatiotemporal trajectory segment of the target vehicle based on the vehicle passage records of the target vehicle, and determine the driving characteristics of the target vehicle in the spatiotemporal trajectory or spatiotemporal trajectory segment.

[0056] The module is configured to determine at least the vehicles that meet the preset equivalence conditions based on the vehicle passage records, so as to construct an equivalence reference vehicle group;

[0057] The second determining module is configured to determine the baseline driving characteristics of the peer reference vehicle group;

[0058] The third determining module is configured to determine the difference characteristics of the target vehicle relative to the equivalent reference vehicle group based on the driving characteristics of the target vehicle and the baseline driving characteristics of the equivalent reference vehicle group.

[0059] The evaluation module is configured to input the difference features into a preset driver state evaluation model to output the driver state evaluation result corresponding to the target vehicle.

[0060] In a third aspect, embodiments of the present invention provide a vehicle-road cooperative system, comprising:

[0061] Multiple vehicle-road cooperative communication devices are installed along the road, the vehicle-road cooperative communication devices including a roadside unit (RSU); an on-board unit (OBU) installed in the vehicle; and a driver status assessment system according to the second aspect.

[0062] The driver state assessment method according to embodiments of the present invention can acquire vehicle traffic records collected by multiple vehicle-road cooperative communication devices set up along the road, wherein the vehicle traffic records include at least vehicle identification, travel time, and collection point location information; based on the vehicle traffic records of a target vehicle, a spatiotemporal trajectory or spatiotemporal trajectory segment of the target vehicle is formed, and the driving characteristics of the target vehicle in the spatiotemporal trajectory or spatiotemporal trajectory segment are determined; at least based on the vehicle traffic records, vehicles that meet preset equivalence conditions are determined to construct an equivalence reference vehicle group; the baseline driving characteristics of the equivalence reference vehicle group are determined; based on the driving characteristics of the target vehicle and the baseline driving characteristics of the equivalence reference vehicle group, the difference characteristics of the target vehicle relative to the equivalence reference vehicle group are determined; at least the difference characteristics are input into a preset driver state assessment model to output the driver state assessment result corresponding to the target vehicle. The method of the present invention, by adopting a strategy of dynamically constructing an equivalence reference vehicle group and calculating baseline driving characteristics, can establish a reference benchmark based on other vehicles in the same or similar traffic environment conditions as the target vehicle, thereby enabling the assessment results to dynamically adapt to dynamic road condition changes such as weather and congestion, significantly improving the accuracy and reliability of driver state assessment.

[0063] In a further embodiment of the invention, the difference features also include speed deviation and / or ride comfort anomaly index. In this embodiment, by considering the relative deviation between the target vehicle's speed and the baseline average speed of the equivalent reference vehicle group, and / or the deviation of the target vehicle's speed change rate relative to the equivalent reference vehicle group across multiple consecutive road segments, the driver's state can be comprehensively assessed from two dimensions: speed amplitude deviation and speed change smoothness, resulting in a more comprehensive and accurate assessment.

[0064] In a further embodiment of the invention, traffic flow coordination characteristics are determined based on the dispersion of the baseline driving characteristics of the equivalent reference vehicle group, and these characteristics are input into the evaluation model along with the difference characteristics. This embodiment's approach can give high attention to slight deviations of the target vehicle when the speed distribution of the equivalent reference vehicle group is concentrated (highly coordinated traffic flow); when the speed distribution of the equivalent reference vehicle group is discrete (chaotic traffic flow), the weight of the target vehicle's deviation is appropriately weakened. This allows for dynamic adjustment of the evaluation sensitivity based on the current level of traffic flow coordination, maintaining high accuracy under different traffic conditions.

[0065] Other optional features and technical effects of the embodiments of the present invention are partly described below and partly apparent from reading this document. Attached Figure Description

[0066] Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings. The elements shown are not limited to the scale shown in the drawings, and the same or similar reference numerals in the drawings denote the same or similar elements, wherein:

[0067] Figure 1 An exemplary flowchart of a driver state assessment method according to an embodiment of the present invention is shown;

[0068] Figure 2 An exemplary flowchart of a driver state assessment method according to an embodiment of the present invention is shown;

[0069] Figure 3 An exemplary flowchart of a driver state assessment method according to an embodiment of the present invention is shown;

[0070] Figure 4 An exemplary flowchart of a driver condition assessment system according to an embodiment of the present invention is shown;

[0071] Figure 5 An exemplary flowchart of a driver condition assessment system according to an embodiment of the present invention is shown;

[0072] Figure 6 An exemplary block diagram of a vehicle-road cooperative system according to an embodiment of the present invention is shown; and

[0073] Figure 7 An exemplary structural diagram of an electronic device capable of implementing the method according to an embodiment of the present invention is shown. Detailed Implementation

[0074] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments and accompanying drawings. Here, the illustrative embodiments and descriptions of this invention are used to explain the invention, but are not intended to limit the invention.

[0075] Exemplary embodiments of the invention will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the invention are shown in the drawings, it should be understood that the invention may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this invention will be thorough and complete, and will fully convey the scope of the invention to those skilled in the art.

[0076] The user / vehicle data, data acquisition, and / or use involved in the embodiments of this invention strictly comply with the laws, regulations, and industry standards of relevant countries and regions. The collection and acquisition of data involved in the embodiments of this invention are all done in advance by actively prompting or prominently displaying information to inform users and obtaining authorization, or by obtaining full authorization from all parties. The processing, manipulation, forwarding, and use of data involved in the embodiments of this invention are all carried out on the premise that the user or relevant party is fully informed and authorized. In implementing the various embodiments of this invention, the types of data or information, scope of use, and usage scenarios that may be involved are informed to users or relevant parties and authorization is obtained through appropriate means. The specific methods of notification and authorization may vary according to actual circumstances, and this invention is not limited in this regard. The processing of personal information / vehicle information involved in the embodiments of this invention is carried out under the premise of having a legal basis (such as obtaining the consent of the personal information subject or being necessary for the performance of a contract), and is only processed within the prescribed or agreed scope. Sensitive personal information such as vehicle identification information and precise positioning information involved in the embodiments of this invention are processed under the premise of having a specific purpose and sufficient necessity, and with the separate authorization and consent of the user or relevant party. In some embodiments of this invention, if the user or relevant party refuses to process personal information other than the information necessary for basic functions, it will not affect the use of the basic functions of the embodiments of this invention.

[0077] As mentioned earlier, in the field of road transportation, the driver's driving condition has a significant impact on driving safety. This is because poor driving conditions, such as fatigue, are a major contributing factor to traffic accidents, especially during long-distance driving and nighttime driving, when drivers are more prone to fatigue, thus affecting their perception and response to the road environment.

[0078] Several driver monitoring solutions have been proposed to address this issue. For example, some known solutions use in-vehicle cameras to capture facial images of the driver and analyze features such as eyelid closure frequency, blinking frequency, and head posture using facial recognition technology to determine fatigue levels. Additionally, some known solutions use wearable devices or in-vehicle sensors to collect physiological signals such as heart rate, electroencephalogram (EEG), and skin conductance to assess driver condition. However, this invention recognizes that while these solutions can directly obtain the driver's physiological or behavioral characteristics, their implementation relies on the collection, transmission, and storage of sensitive personal information such as facial images and physiological signals. Because these solutions involve personal privacy data, they pose risks of data security and privacy breaches. Furthermore, drivers may resist continuous monitoring of their facial or physiological state, affecting the acceptance and widespread application of these solutions.

[0079] Furthermore, this invention recognizes that known solutions based on image recognition or physiological signal monitoring typically require the installation of driver-facing cameras, infrared sensors, physiological monitoring equipment, or wearable devices within the vehicle's cab. These solutions impose high hardware requirements on the vehicles themselves, necessitating the modification or retrofitting of a large number of existing vehicles, resulting in high modification costs and long implementation cycles. Moreover, differences in installation conditions and interface standards across different vehicle models make standardized deployment difficult. These factors hinder the rapid rollout and application of these known solutions within existing road networks.

[0080] Furthermore, some known driver condition assessment schemes use fixed thresholds for judgment. Specifically, some known schemes compare driving time with a preset time threshold, and determine fatigue driving when the driving time exceeds the threshold; other schemes compare vehicle speed with a preset speed threshold. However, this invention recognizes that such fixed threshold schemes have significant limitations: on the one hand, different drivers have individual differences in fatigue tolerance, and the same driving time may have different effects on different drivers in different situations; on the other hand, fixed thresholds are difficult to reflect real-time changes in the traffic environment. For example, in rainy or snowy weather or congested road sections, it is reasonable for normal drivers to reduce their overall driving speed. If a fixed speed threshold is still used for judgment in this case, false alarms are likely to occur, which seriously reduces the accuracy and reliability of the assessment.

[0081] To overcome the limitations of fixed thresholds, some known solutions use historical averages as a reference, such as comparing the current speed of a target vehicle with the historical average speed of that road segment. However, this invention recognizes that such historical average speeds only reflect the statistical characteristics of all past times and are insufficient to reflect the actual traffic conditions at the current moment. For the purpose of explanation, and not limitation, when the overall real-time traffic speed decreases due to temporary construction, traffic accidents, or sudden weather changes on a road segment, judgments based on historical average speeds may misinterpret normal adaptive deceleration as abnormal driving. In other words, while historical average solutions are an improvement over fixed thresholds, they still lack the ability to dynamically adapt to current traffic conditions.

[0082] Furthermore, some known solutions employ video analytics to construct a reference baseline by simultaneously detecting a group of vehicles from the same detection unit (such as a roadside camera), comparing the target vehicle's driving status with surrounding vehicles. However, this invention further recognizes that such solutions have fundamental limitations in scenarios with sparse traffic, such as late-night hours on highways: when a target vehicle passes a detection point, there may be no other vehicles simultaneously within the detection range for an extended period, making it impossible to construct an effective reference baseline and causing the solution to fail. In other words, solutions relying on "simultaneously detected groups of vehicles within the same field of view" cannot achieve continuous evaluation capabilities across all weather conditions and road sections.

[0083] To address the technical problems existing in the aforementioned known solutions, embodiments of the present invention provide a driver state assessment method and a corresponding driver state assessment system and vehicle-road cooperative system, which can accurately and effectively assess driver state. Several embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0084] In an embodiment of the present invention, reference is made to Figure 1 A method for assessing driver condition is provided, which may include the following steps S110, S120, S130, S140, S150 and S160.

[0085] S110: Obtain vehicle passage records collected by multiple vehicle-road cooperative communication devices set up along the road.

[0086] In some embodiments, the vehicle-road cooperative communication device includes a device deployed along a road that can communicate with on-board units and collect vehicle traffic information.

[0087] In some embodiments, the vehicle-to-infrastructure (V2I) communication device may include a Roadside Unit (RSU). By way of explanation and not limitation, the Roadside Unit (RSU) can communicate with the Onboard Unit (OBU) via V2I communication protocols such as DSRC or C-V2X to obtain relevant vehicle information as the vehicle passes. In some embodiments, the V2I communication device may include an ETC gantry system, which can obtain vehicle passage records through interaction with the onboard ETC tag as the vehicle passes.

[0088] In some embodiments, vehicle-to-infrastructure (V2I) communication devices may be deployed in a distributed manner along the road (extension) direction. In this embodiment, there may be a predetermined spacing between adjacent devices. By way of explanation and not limitation, in highway scenarios, the spacing between adjacent V2I communication devices may include, but is not limited to, several kilometers to tens of kilometers. This distributed deployment method enables V2I communication devices to collect traffic information at multiple locations during vehicle travel, thereby constructing the spatiotemporal trajectory of the vehicle.

[0089] In some embodiments, vehicle passage records include information collected and recorded by vehicle-to-infrastructure (V2I) communication devices when a vehicle passes by. In some embodiments, vehicle passage records may include at least vehicle identification, passage time, and collection point location information.

[0090] In some embodiments, a vehicle identifier may be used to uniquely identify a vehicle. In some embodiments, the vehicle identifier includes, but is not limited to, an encrypted license plate number, an ETC card number, an on-board unit number, etc.

[0091] In some embodiments, the passage time can be used to record the moment a vehicle passes through a data collection point or the time interval between two data collection points. In some embodiments, the passage time may include a timestamp accurate to the second or millisecond.

[0092] In some embodiments, the location information of the collection point can be used to identify the location traversed by the vehicle. In some embodiments, the location information of the collection point may include the geographical coordinates, road number, and station number of the vehicle-road cooperative communication device.

[0093] In some embodiments, vehicle passage records may further include vehicle type information. In some embodiments, vehicle type information may include vehicle category, model code, etc. In one embodiment, vehicle category may include, but is not limited to, passenger cars, trucks, special-purpose vehicles, trailers, and vehicle-to-everything (V2X) trains. In some embodiments, vehicle type information may be obtained from vehicle registration information stored in the on-board unit, or obtained through the identification function of the vehicle-to-infrastructure (V2I) communication device.

[0094] In some embodiments, step S110 may be performed in response to the time when the vehicle arrives at the current collection point.

[0095] S120: Based on the vehicle passage records of the target vehicle, form the spatiotemporal trajectory or spatiotemporal trajectory segment of the target vehicle, and determine the driving characteristics of the target vehicle in the spatiotemporal trajectory or spatiotemporal trajectory segment.

[0096] In some embodiments, the target vehicle includes the vehicle for which driver status assessment is currently required. In some embodiments, the vehicle may be identified as the target vehicle if its passage record is updated. In some embodiments, the vehicle may be identified as the target vehicle if it passes through the data collection point.

[0097] In some embodiments, a spatiotemporal trajectory may include the travel path records of the target vehicle in both time and space dimensions. In some embodiments, a spatiotemporal trajectory may be formed by sorting and associating multiple passage records of the target vehicle according to passage time. In some embodiments, a spatiotemporal trajectory may include a complete spatiotemporal trajectory covering multiple road segments, or it may include spatiotemporal trajectory segments containing multiple data collection point nodes.

[0098] In some embodiments, the spatiotemporal trajectory can be obtained by sorting and associating multiple vehicle passage records generated by the target vehicle at roadside collection points according to passage time. In some embodiments, each vehicle passage record can correspond to a trajectory point in the spatiotemporal trajectory. In some embodiments, the spatiotemporal trajectory can include the driving time-series data (discrete time series) of the target vehicle.

[0099] In some embodiments, the spatiotemporal trajectory can determine the interval travel time, interval average speed, speed dispersion, and temporal features such as speed sequences / trends obtained from multiple consecutive intervals between adjacent or non-adjacent collection points, thereby providing input for subsequent differential feature extraction and driver state assessment. By way of explanation and not limitation, unlike methods that only utilize segment-level summary statistics (e.g., historical average speed, instantaneous speed obtained from a single detection, or comparison with a fixed threshold), this invention employs association information including collection point locations and corresponding timestamps during the trajectory construction stage, forming a trajectory point sequence in chronological order, thus being more suitable for online driver state assessment and adaptive real-time road condition changes.

[0100] In some embodiments, the step of forming a spatiotemporal trajectory or spatiotemporal trajectory segment of the target vehicle based on the vehicle passage records of the target vehicle may include the following steps A1 to A3. A1: Filtering target passage records belonging to the target vehicle from the vehicle passage record set according to the vehicle identifier; A2: Sort the filtered target passage records by passage time; A3: Associating the sorted passage records to form a spatiotemporal trajectory or spatiotemporal trajectory segment containing multiple collection point nodes.

[0101] In some embodiments, driving features (parameters) may include parameters extracted from spatiotemporal trajectories or segments of spatiotemporal trajectories that reflect the vehicle's driving state. In some embodiments, driving features may include data collection time, average speed over a given interval, travel time over a given interval, and rate of change of speed. In some embodiments, the average speed over a given interval may be calculated based on the distance between adjacent data collection points and the time difference between the time the target vehicle travels through adjacent data collection points.

[0102] In some embodiments, when the vehicle-road cooperative communication device has speed measurement capabilities, the driving characteristics may also include the instantaneous speed of the vehicle passing the collection point.

[0103] S130: At least vehicles that meet the preset equivalence conditions are identified based on vehicle passage records in order to construct an equivalence reference vehicle group.

[0104] In some embodiments, the peer reference vehicle group includes a group of vehicles that meet preset peer conditions with the target vehicle.

[0105] In some embodiments, preset peer conditions may include conditions for selecting peer reference vehicles. In this embodiment, preset peer conditions may be used to constrain the peer reference vehicles and the target vehicle to be comparable in traffic environment dimension and / or vehicle attribute dimension, so that the driving characteristics of the peer reference vehicle group can serve as a reference benchmark for evaluating the driving status of the target vehicle.

[0106] In some embodiments, the preset equivalence conditions may include one or more dimensions. In this embodiment i, the preset equivalence conditions may include, but are not limited to, one or more of the following dimensions: vehicle type, space, direction, and time.

[0107] In some embodiments, the step of determining vehicles that meet preset equivalence conditions to construct an equivalence reference vehicle group may include: determining vehicles that have a spatial association with the target vehicle based on the collection point location information in the vehicle passage records.

[0108] In some embodiments, vehicles spatially associated with the target vehicle may include vehicles that have passed through the current acquisition point in the spatiotemporal trajectory of the target vehicle. By way of explanation and not limitation, the vehicles in the peer reference vehicle group determined through spatial association in embodiments of the present invention pass through the same acquisition point as the target vehicle. Therefore, the vehicles in the peer reference vehicle group and the target vehicle face the same road geometry conditions (such as slope, curves, number of lanes, etc.) and road infrastructure conditions (such as speed limit signs, lighting conditions, etc.), thereby making the baseline driving characteristics calculated based on this vehicle group spatially comparable.

[0109] In some embodiments, the step of determining vehicles spatially associated with the target vehicle based on the location information of the collection point in the vehicle passage record may include: determining vehicles spatially associated with the target vehicle from vehicles within a preset time window based on the passage time in the vehicle passage record. In this embodiment, the preset time window is determined, for example, based on the passage time of the target vehicle passing through the current collection point.

[0110] In some embodiments, vehicles spatiotemporally associated with the target vehicle may include vehicles that pass through the current data collection point within a preset time window. As an explanation, and not a limitation, some known schemes use historical average speed as a baseline. However, historical average speed cannot reflect real-time factors such as current weather conditions, traffic congestion, and road construction. This coarse-grained driver state assessment scheme based on a fixed historical baseline cannot ensure that the baseline driving characteristics reflect the current real-time traffic conditions. Therefore, this embodiment of the invention, through the limitation of a time window, ensures that vehicles in the determined equivalent reference vehicle group and the target vehicle pass through the same data collection point within a similar timeframe. This results in vehicles in the reference vehicle group and the target vehicle facing similar traffic flow conditions, thereby making the baseline driving characteristics timely and dynamically adaptable to the current traffic environment.

[0111] In an optional embodiment, the peer reference vehicle group may include at least some vehicles that are not within the detection range of the vehicle-road cooperative communication equipment corresponding to the current collection point and / or the collection point adjacent to the current collection point when the target vehicle passes the current collection point, and are detected within a preset time window.

[0112] As an explanation, and not a limitation, some known solutions use a group of vehicles detected simultaneously by the same detection unit (such as a camera or radar) to construct a baseline. However, this invention recognizes that this approach has insurmountable drawbacks in sparse traffic scenarios, such as late nights on highways, where no other vehicles may be within the detection range for an extended period, making it impossible to construct an effective baseline and leading to evaluation failure. To address this, unlike known technologies that can only use "vehicles detectable at this moment," this invention employs a time window accumulation mechanism: using "vehicles passing through the time window." These vehicles include, but are not limited to, those that passed the collection point before the target vehicle arrived and are currently ahead; and those that arrived at the collection point after the target vehicle passed and are currently behind. These vehicles are not within the detection range at the time the target vehicle is detected, but can still serve as members of the peer reference vehicle group. Therefore, the method of this invention can still construct a statistically significant peer reference vehicle group in sparse traffic scenarios, ensuring the reliability and continuity of the evaluation and achieving all-weather, all-road segment driver status evaluation capabilities.

[0113] In some embodiments, the peer reference vehicle group may include at least some vehicles that are not located within the collection area of ​​the collection point at the time the target vehicle is detected, and that pass through the collection point within a preset time window.

[0114] In some embodiments, the peer reference vehicle group may include at least some vehicles that are not within the detection field of view of the vehicle-road cooperative communication device at the time the target vehicle is detected, and that are detected within a preset time window.

[0115] In some embodiments, the peer reference vehicle group may include at least some vehicles that are not detectable by the vehicle-road cooperative communication device at the time the target vehicle is detected, and that are detected within a preset time window.

[0116] In some embodiments, the peer reference vehicle group may include at least some vehicles that are not located near the current collection point when the target vehicle passes the current collection point, and that pass the current collection point within a preset time window.

[0117] In some embodiments, the peer reference vehicle group may include at least some vehicles that do not pass through the current collection point at the same time as the target vehicle, but pass through the current collection point within a preset time window.

[0118] In an optional embodiment, the step of determining vehicles spatially associated with the target vehicle based on the location information of the collection points in the vehicle passage record may include: determining vehicles with the same travel direction as the target vehicle from among vehicles that have passed through the same road segment as the target vehicle, based on the location information of the collection points.

[0119] In some embodiments, the same road segment includes the road segment between at least two adjacent collection points traversed by the target vehicle. As an explanation, and not a limitation, the above embodiments of the present invention further enhance spatial comparability by simultaneously defining the same road segment and the same direction of travel. Specifically, road conditions in different directions on a two-way road may differ (e.g., positive or negative slope, location of entrances and exits, etc.), and defining the same direction of travel can eliminate the impact of such differences.

[0120] In an optional embodiment, the step of determining vehicles that meet preset equivalence conditions may further include: determining vehicles belonging to the same vehicle type as the target vehicle from among vehicles that have a spatiotemporal association with the target vehicle, based on vehicle attribute information in the vehicle passage record. In this embodiment, the vehicle type is determined, for example, based on vehicle model information in the vehicle passage record. As an explanation, and not a limitation, different types of vehicles have significant differences; for example, the normal driving speeds of passenger cars and heavy trucks differ significantly. If both are mixed to calculate the baseline, the reference value of the baseline will be significantly reduced. By limiting the vehicle type to the same type, this embodiment of the invention ensures that vehicles in the equivalent reference vehicle group have similar power performance, thereby making the baseline driving characteristics more accurate.

[0121] In a further embodiment, the preset equivalence conditions may include: being the same or similar vehicle type as the target vehicle, on the same road segment, traveling in the same direction, and passing through the same data collection point within a preset time window. As an explanation, by combining the above-mentioned multi-dimensional equivalence conditions, it can be ensured that vehicles in the equivalence reference vehicle group face highly similar traffic environments to the target vehicle, thereby making the baseline driving characteristics calculated based on the equivalence reference vehicle group have high reference value.

[0122] In an optional embodiment, the step of determining vehicles that meet preset equivalence conditions may further include: based on the load information in the vehicle passage record, determining vehicles belonging to the same load class as the target vehicle from among vehicles of the same vehicle type. As an explanation, and not a limitation, even among trucks, the speed characteristics of empty and fully loaded trucks may differ significantly. This embodiment of the invention, through further screening by load class, can make the baseline driving characteristics more accurate, thereby further improving the accuracy of the assessment.

[0123] In an optional embodiment, the step of determining vehicles spatially associated with the target vehicle from vehicles within a preset time window may include: determining vehicles spatially associated with the target vehicle from vehicles that passed through the current collection point within a preset time period before the target vehicle passed through the current collection point. In one specific embodiment, the preset time window is 5 to 15 minutes before the target vehicle passes through the current collection point. In an optional embodiment, the preset time window may also include a period of time after the target vehicle passes through the current collection point, for example, 10 minutes before and 5 minutes after.

[0124] In an optional embodiment, the duration of the preset time window can be dynamically adjusted according to the number of vehicles that meet the preset equivalence conditions within the preset time window.

[0125] In some embodiments, the duration of the preset time window can be extended when the number of vehicles is less than a preset minimum sample size. As an explanation, and not a limitation, the present invention recognizes that in scenarios with significant traffic flow variations, the equivalent reference vehicle group may struggle to obtain a sufficient sample size under certain low traffic conditions. To address this, embodiments of the present invention dynamically adjust the duration of the time window, automatically extending the time window to accumulate more samples when traffic flow is sparse, and using a shorter time window to ensure timeliness when traffic flow is dense.

[0126] Accordingly, in some embodiments, the step of determining vehicles that meet preset equivalence conditions to construct an equivalence reference vehicle group may include: filtering vehicles that meet preset equivalence conditions from vehicle passage records within a preset time window; when the number of vehicles that meet preset equivalence conditions is less than a preset threshold, expanding the preset time window to obtain an expanded time window; and filtering vehicles that meet preset equivalence conditions from vehicle passage records within the expanded time window to construct an equivalence reference vehicle group.

[0127] In an optional embodiment, the step of determining vehicles spatially associated with the target vehicle based on the location information of the collection points in the vehicle passage record may include: determining vehicles spatially associated with the target vehicle from vehicles passing through the collection points within a preset time window; when the number of vehicles spatially associated with the target vehicle is less than a preset threshold, expanding the preset time window to obtain an extended time window; determining vehicles spatially associated with the target vehicle from vehicles passing through the collection points within the extended time window to construct a peer reference vehicle group.

[0128] In an optional embodiment, the step of determining vehicles spatially associated with the target vehicle may include: selecting a preset number of vehicles that most recently meet preset equivalence conditions to construct an equivalence reference vehicle group. In a specific embodiment, the 100 vehicles that most recently passed through the current collection point and meet the equivalence conditions are selected. As an explanation, and not a limitation, the above embodiments use a fixed sample size rather than a fixed time window to construct the equivalence reference vehicle group, thereby ensuring the stability of the sample size in scenarios with large fluctuations in traffic flow.

[0129] S140: Determine the baseline driving characteristics of the equivalent reference vehicle group.

[0130] In some embodiments, the baseline driving characteristics include statistical values ​​of the driving characteristics of an equivalent reference vehicle group, and the baseline driving characteristics can serve as a reference benchmark for evaluating the driving status of the target vehicle.

[0131] In some embodiments, the step of determining the baseline driving characteristics of the equivalent reference vehicle group may include the following steps B1 and B2. B1 determines the driving characteristics of each vehicle in the equivalent reference vehicle group; B2 performs statistical processing on the driving characteristics of each vehicle to obtain the baseline driving characteristics. In some embodiments, the driving characteristics may include data acquisition time, average speed over an interval, travel time over an interval, rate of change of speed, etc. In this embodiment, a detailed description of the driving characteristics can be found in the relevant description in step 120 above, and will not be repeated here.

[0132] In some embodiments, baseline driving characteristics may include baseline average speed and speed dispersion index. In some embodiments, baseline average speed may include the arithmetic mean of the interval average speeds of each vehicle in the equivalent reference vehicle group. In some embodiments, speed dispersion index may include the standard deviation of the interval average speeds of each vehicle in the equivalent reference vehicle group.

[0133] In some embodiments, baseline driving characteristics can be updated in real time based on vehicle passage records of newly added vehicles in the peer reference vehicle group, thereby maintaining the timeliness of the baseline.

[0134] In some embodiments, outlier filtering may be performed on the driving characteristics of each vehicle when calculating baseline driving characteristics. In some embodiments, outlier filtering may include one or more of the following processes: using the median instead of the arithmetic mean, calculating a cutoff mean after removing the highest and lowest percentages of samples, and iteratively removing samples that deviate from the initial baseline by more than a preset multiple of standard deviation. By way of explanation and not limitation, outlier filtering may improve the stability of the baseline.

[0135] In this embodiment of the invention, the driving characteristics of the target vehicle can be determined based on two or more collection points it passes through (i.e., two or more consecutive road segments) and the travel time between the two or more collection points it passes through.

[0136] Accordingly, in some embodiments, reference is made to Figure 2 The step of forming a spatiotemporal trajectory or spatiotemporal trajectory segment of the target vehicle based on the vehicle passage record of the target vehicle, and determining the driving characteristics of the target vehicle in the spatiotemporal trajectory or spatiotemporal trajectory segment may include the following step S210.

[0137] S210: Calculate the first driving characteristics of the target vehicle based on the distance between at least two collection points and the travel time of the target vehicle through at least two collection points.

[0138] In some embodiments, the first driving feature may include driving features calculated based on traffic data of the target vehicle between two adjacent collection points. In some embodiments, the two collection points include at least two collection points selected sequentially from a plurality of collection points along the road.

[0139] In some embodiments, the first driving characteristic may include interval driving time. In this embodiment, interval driving time is, for example, the difference between the timestamps of the target vehicle passing through two adjacent collection points.

[0140] In some embodiments, the first driving characteristic may include the average vehicle speed over an interval. In some embodiments, determining the average vehicle speed over an interval may include the following steps C1 to C4: C1: Obtain the first travel time of the target vehicle passing through the first collection point and the second travel time passing through the second collection point; C2: Obtain the first geographical location information corresponding to the first collection point and the second geographical location information corresponding to the second collection point; C3: Determine the interval distance between the first collection point and the second collection point based on the first geographical location information and the second geographical location information; C4: Calculate the average vehicle speed over an interval based on the first travel time, the second travel time, and the interval distance.

[0141] In some embodiments, such as after step S210 described above, the step of determining the baseline driving characteristics of the equivalent reference vehicle group may include steps S220 and S230 described below.

[0142] S220: Calculate the first reference driving characteristics of each peer reference vehicle based on the distance between at least two collection points and the travel time of each peer reference vehicle between at least two collection points.

[0143] In some embodiments, the first reference driving characteristic may include the first driving characteristics of each vehicle in the equivalent reference vehicle group. In this embodiment, the specific calculation method of the first reference driving characteristic can refer to the aforementioned calculation method of the first driving characteristic, the difference being that the calculation object is the equivalent reference vehicle rather than the target vehicle, and will not be repeated here.

[0144] S230: Statistical processing is performed on the first reference driving characteristics to obtain the first baseline driving characteristics.

[0145] In some embodiments, the first baseline driving characteristic may include statistical values ​​of a first reference driving characteristic of an equivalent reference vehicle group. In some embodiments, statistical processing may include, but is not limited to, calculating the arithmetic mean, median, standard deviation, etc.

[0146] In some embodiments, the first baseline driving characteristic may include the baseline average vehicle speed and the baseline speed standard deviation. In this embodiment, the baseline average vehicle speed may include the arithmetic mean of the interval average vehicle speeds of each vehicle in the equivalent reference vehicle group. In some embodiments, the baseline speed standard deviation may include the standard deviation of the interval average vehicle speeds of each vehicle in the equivalent reference vehicle group. For explanation, the baseline speed standard deviation can be used to characterize the degree of dispersion of the speed distribution within the equivalent reference vehicle group.

[0147] In some embodiments, unlike the aforementioned single-segment-based scheme, calculations can also be performed based on traffic data from at least three collection points (i.e., at least two consecutive road segments) to obtain statistical information on the driving characteristics of the target vehicle on multiple consecutive road segments.

[0148] In some embodiments, reference Figure 3 The steps of forming a spatiotemporal trajectory or spatiotemporal trajectory segment of the target vehicle based on the vehicle passage record of the target vehicle, and determining the driving characteristics of the target vehicle in the spatiotemporal trajectory or spatiotemporal trajectory segment may include the following steps S310, S320 and S330.

[0149] S310: Calculate multiple first driving characteristics of the target vehicle based on the distance between at least three collection points and the travel time of the target vehicle through at least three collection points.

[0150] In some embodiments, when the target vehicle passes through at least three collection points, the first driving characteristics of at least two road segments can be calculated. In one example, when the target vehicle passes through collection points P1, P2, and P3 in sequence, the first driving characteristics of the road segment from P1 to P2 and the first driving characteristics of the road segment from P2 to P3 can be calculated respectively, thereby obtaining multiple first driving characteristics.

[0151] In some embodiments, the multiple first driving characteristics may include the average speed of the target vehicle across various road segments, thereby forming a speed sequence. In this embodiment, the speed sequence can be used to reflect the speed changes of the target vehicle across multiple consecutive road segments.

[0152] S320: Statistical driving features are obtained by statistically processing multiple first driving features.

[0153] In some embodiments, such as in step S320 above, the statistical driving characteristics may include feature values ​​obtained by statistical analysis of multiple first driving characteristics of the target vehicle. In some embodiments, the statistical driving characteristics may include the average and / or standard deviation of multiple first driving characteristics.

[0154] S330: Determine the second driving characteristics of the target vehicle based on multiple first driving characteristics and statistical driving characteristics.

[0155] In some embodiments, the second driving characteristic may include a characteristic determined based on multi-segment data that reflects the behavioral characteristics of the target vehicle during continuous driving.

[0156] In some embodiments, the second driving characteristic may include driving smoothness. In this embodiment, driving smoothness can be used to characterize the rate of change of vehicle speed across multiple consecutive road segments. By way of explanation and not limitation, the driving smoothness of the embodiments of the present invention may reflect the stability of the driver's control over vehicle speed. For example, a driver under normal conditions can usually maintain a relatively stable vehicle speed, while a driver under fatigue may experience increased speed fluctuations.

[0157]

[0158] in, For the driving smoothness of the target vehicle. Let be the average speed of the target vehicle in the j-th target road segment. For the target vehicle in Average vehicle speed on each road segment This refers to the number of road segments.

[0159] S340: Calculate multiple first reference driving characteristics of each peer reference vehicle based on the distance between at least three collection points and the travel time of each peer reference vehicle through at least three collection points.

[0160] In some embodiments, the first reference driving feature may include driving features calculated based on the passage data of each peer reference vehicle in the peer reference vehicle group between two adjacent collection points.

[0161] In some embodiments, the first reference driving characteristic may include the interval driving time. In this embodiment, the interval driving time is, for example, the difference in the timestamps of each pair of reference vehicles passing through two adjacent collection points. In this embodiment, the specific processing logic for the first reference driving characteristic can be referred to the specific description of the first driving characteristic in the aforementioned step S310, the difference being that the processing object is each vehicle in the pair of reference vehicles, which will not be repeated here.

[0162] S350: Statistical processing is performed on multiple first reference driving characteristics of each equivalent reference vehicle to obtain the statistical reference driving characteristics of each equivalent reference vehicle.

[0163] In some embodiments, such as in step S350 above, the statistical reference driving characteristics may include feature values ​​obtained by statistically analyzing multiple first reference driving characteristics of each peer reference vehicle. In some embodiments, the statistical reference driving characteristics may include the average value of multiple first reference driving characteristics, such as average vehicle speed.

[0164] S360: Determine the second reference driving characteristics of each equivalent reference vehicle based on multiple first reference driving characteristics and statistical reference driving characteristics.

[0165] In some embodiments, the second row of reference driving features may include features determined based on multi-segment data that reflect the behavioral characteristics of each peer reference vehicle during continuous driving.

[0166] In some embodiments, the second row of reference driving features may include driving smoothness. In this embodiment, a detailed description of driving smoothness can be found in the relevant description in the aforementioned S330, and will not be repeated here.

[0167] In some embodiments, the driving smoothness of each equivalent reference vehicle can be determined by the following formula:

[0168]

[0169] in, For the driving smoothness of each equivalent reference vehicle, Let be the interval average speed of each equivalent reference vehicle on the i-th target road segment. For each equivalent reference vehicle in Average vehicle speed on each road segment This refers to the number of road segments.

[0170] S370: Statistical processing is performed on the second reference driving characteristics to obtain the second baseline driving characteristics.

[0171] In some embodiments, the second baseline driving characteristic may include statistical values ​​of second reference driving characteristics of a group of equivalent reference vehicles. In some embodiments, the second baseline driving characteristic may include baseline driving comfort. In this embodiment, the baseline driving comfort includes statistical values ​​(e.g., the arithmetic mean) of the driving comfort of each equivalent reference vehicle, which can be determined by the following formula:

[0172]

[0173] in, Baseline driving smoothness, For the first The driving smoothness of a comparable reference vehicle. The number of vehicles in the equivalent reference vehicle group.

[0174] In the above embodiments of the present invention, the second driving characteristics (such as driving smoothness) of the target vehicle and the second baseline driving characteristics (such as baseline driving smoothness) of the equivalent reference vehicle group determined based on the traffic data of multiple consecutive road segments can provide a basis for subsequent differential characteristic calculations.

[0175] S150: Based on the driving characteristics of the target vehicle and the baseline driving characteristics of the equivalent reference vehicle group, determine the difference characteristics of the target vehicle relative to the equivalent reference vehicle group.

[0176] In this embodiment of the invention, the difference feature includes a quantitative representation of the degree of deviation of the target vehicle's driving characteristics from the baseline driving characteristics of a peer reference vehicle group. In this embodiment, the difference feature can be used to reflect the difference between the target vehicle's driving state in the current traffic environment and that of other vehicles under the same conditions.

[0177] In some embodiments, the difference feature may include at least one of speed deviation feature and time deviation feature. In some embodiments, the difference feature may include the standardized deviation of the target vehicle's driving characteristics from the corresponding parameters in the baseline driving characteristics.

[0178] In some embodiments, the differential characteristics may include a sequence of deviations of the target vehicle across multiple road segments in its spatiotemporal trajectory, and a comprehensive deviation index determined based on the deviation sequence. As an explanation, and not a limitation, analyzing the trend of deviation changes of the target vehicle across multiple consecutive road segments can more accurately determine changes in the driver's state.

[0179] In some embodiments, the first driving characteristic may include vehicle speed. In this embodiment, the (first) difference characteristic may include vehicle speed deviation.

[0180] In some embodiments, the vehicle speed deviation can be used to characterize the relative deviation between the speed of the target vehicle and the average speed of the equivalent reference group of vehicles.

[0181] In some embodiments, the vehicle speed deviation can be determined by the following formula:

[0182]

[0183] in, For vehicle speed deviation, The average speed of the target vehicle between two adjacent data collection points. This represents the baseline average speed of the equivalent reference vehicle group on the same road segment. By way of interpretation and not limitation, speed deviation reflects the degree to which the target vehicle's speed deviates from the average level of the current traffic flow. Specifically, a high speed deviation indicates a significant difference between the target vehicle's speed and the speeds of other vehicles under similar conditions, suggesting a risk of driver fatigue or other abnormal conditions.

[0184] In some embodiments, a target vehicle may be identified as having an abnormal speed if the speed deviation exceeds a preset deviation threshold. In a specific example, if the preset deviation threshold is 30%, then the target vehicle is marked as having an abnormal speed when the speed deviation exceeds 30%.

[0185] In some embodiments, the second driving characteristic includes driving smoothness. In this embodiment, driving smoothness can be used to characterize the rate of change of vehicle speed across multiple consecutive road segments.

[0186] In some embodiments, the driving smoothness of the target vehicle can be determined by the following formula:

[0187]

[0188] in, For the driving smoothness of the target vehicle. Let be the average speed of the target vehicle in the j-th target road segment. For the target vehicle in Average vehicle speed on each road segment This refers to the number of road segments.

[0189] In some embodiments, the driving smoothness of each equivalent reference vehicle can be determined by the following formula:

[0190]

[0191] in, For the driving smoothness of each equivalent reference vehicle, For each equivalent reference vehicle in the first The average speed of vehicles on each target road segment. For the target vehicle in Average vehicle speed on each road segment This refers to the number of road segments.

[0192] In some embodiments, the (second) difference feature may include a ride comfort anomaly index. In this embodiment, the ride comfort anomaly index can be used to characterize the degree of deviation of the target vehicle's driving ride comfort from the baseline driving ride comfort of an equivalent reference vehicle group.

[0193] In some embodiments, the smoothness anomaly index can be determined by the following formula:

[0194]

[0195] in, The smoothness anomaly index for the target vehicle. For the driving smoothness of the target vehicle. For the driving smoothness of each equivalent reference vehicle The average value.

[0196] As an explanation, and not a limitation, the ride comfort anomaly index reflects the difference in speed fluctuation of a target vehicle relative to other vehicles under the same conditions. Specifically, when the ride comfort anomaly index is significantly greater than 1, it indicates that the speed fluctuation of the target vehicle is greater than the average level of the equivalent reference vehicle group, which can be inferred to be a decrease in the driver's ability to control the vehicle speed, thus posing a risk of driver fatigue.

[0197] In some embodiments, if the smoothness anomaly index exceeds a preset smoothness threshold, it can be determined that the target vehicle has a driving smoothness anomaly.

[0198] In some embodiments, outlier filtering may be performed on the driving comfort of each peer reference vehicle before calculating the baseline driving comfort. This outlier filtering eliminates the influence of extreme individuals on the baseline, thereby improving the stability and representativeness of the baseline.

[0199] In some embodiments, the difference features may simultaneously include speed deviation (first difference feature) and ride comfort anomaly index (second difference feature). In this embodiment, the driving state of the target vehicle is comprehensively determined by combining two dimensions: speed magnitude deviation and speed change smoothness. By way of explanation and not limitation, speed deviation can be used to measure the difference between the target vehicle's speed and the baseline; ride comfort anomaly index can be used to measure whether the speed fluctuation of the target vehicle is abnormal. Combining the two allows for a more comprehensive assessment of the driver's state.

[0200] In an optional embodiment, the difference feature may further include a speed trend deviation feature. In this embodiment, the speed trend deviation feature can be used to characterize the degree of consistency between the target vehicle and the equivalent reference vehicle group in the direction of speed change. As an explanation, and not a limitation, when the equivalent reference vehicle group as a whole is decelerating (e.g., due to congestion or construction ahead) while the target vehicle does not decelerate accordingly, it can be inferred that the driver's perception of the road conditions ahead is reduced and there is a higher safety risk.

[0201] In some embodiments, the speed trend deviation feature can be determined based on the sign consistency rate between the speed change direction of the target vehicle between multiple consecutive collection points and the speed change direction of the equivalent reference vehicle group between corresponding collection points.

[0202] S160: At least input the difference features into the preset driver state assessment model to output the driver state assessment result corresponding to the target vehicle.

[0203] In some embodiments, the driver state assessment model may include a pre-trained model. In this embodiment, the driver state assessment model may establish a mapping relationship between differential features and driver state assessment results. In some embodiments, the driver state assessment model may output a driver state category, state score, and / or risk probability based on the input differential features.

[0204] In some embodiments, the driver state assessment model includes a model trained through steps D1 to D4 described below.

[0205] D1: Retrieves event logs from external annotation data sources.

[0206] In some embodiments, the external annotation data source includes data sources capable of providing accurate annotations of driver status. In some embodiments, the external annotation data source may include traffic accident data and / or fleet monitoring data equipped with onboard monitoring devices. By way of explanation and not limitation, traffic accident data may come from the accident recording system of a traffic management department, wherein accidents caused by fatigue driving can serve as a positive sample source of fatigue status; fleet monitoring data may come from transport fleets equipped with driver monitoring devices (such as fatigue detection cameras, heart rate monitoring devices, etc.), which can provide real-time annotations of driver status.

[0207] D2: Spatiotemporally correlate event records with vehicle passage records to determine the vehicle passage record corresponding to the event record.

[0208] In some embodiments, spatiotemporal correlation includes matching vehicle passage records within a corresponding spatiotemporal range based on the time and location of the event. As an explanation, and not a limitation, for a fatigue driving accident, the passage records of the vehicle at various collection points within a certain period prior to the accident can be traced back based on the time and location of the accident.

[0209] D3: Extract sample features from the associated vehicle passage records and label the corresponding driver status to establish a labeled dataset.

[0210] In some embodiments, the method for extracting sample features can refer to the description in steps S120 to S150 above. In some embodiments, sample features may include, but are not limited to, driving features, baseline driving features, and difference features. In some embodiments, the labeling of driver status can be determined based on the event type of the external labeling data source. In a specific embodiment, the traffic record corresponding to a fatigue driving accident can be labeled as a fatigued state; the traffic record corresponding to normal driving (no accident, no fatigue mark) can be labeled as a normal state.

[0211] D4: A driver state assessment model is trained using a labeled dataset.

[0212] In some embodiments, multidimensional difference features within the same road segment window or the same time window can be spliced ​​together to form a fixed-dimensional input vector, and this vector can be input into the evaluation model to obtain the evaluation result.

[0213] In some embodiments, the driver state assessment model may include a rule-based threshold determination model. In this embodiment, differential features can be compared or logically processed with preset thresholds, threshold ranges, or combinations of rules to determine the driver state assessment result. In one example, determination may be based on single-indicator threshold rules or multi-indicator combined thresholds and logical combination rules.

[0214] In some embodiments, the driver state assessment model may include a supervised learning-based machine learning model. In this embodiment, differential features may be used as vector inputs, and the driver state assessment model obtains model parameters through pre-training to output driver state category, state score, and / or risk probability.

[0215] In some embodiments, the driver state assessment model may be implemented using a weighted linear function, logistic regression, or softmax regression. In some embodiments, the weighted linear function may calculate a score based at least on the differential feature vector, and the resulting score may be used as a risk score or further mapped to a state category. In some embodiments, logistic regression or softmax regression may perform a sigmoid or softmax transformation on the linear output to obtain the probability distribution of each state category, facilitating the use of risk probability and confidence level for subsequent intervention triggering.

[0216] In some embodiments, the driver state assessment model may be implemented using Gradient Boosting Decision Tree (GBDT). For explanation, GBDT is suitable for handling differential features of tabular inputs and can automatically learn nonlinear relationships and feature interactions. Specifically, GBDT can learn that combinations of large speed deviations, poor ride comfort, and continuously increasing trend deviations correspond to higher risks on specific road segments or time periods. In some embodiments, the output of the GBDT model may include category, probability, or continuous risk scores, and may output feature importance.

[0217] In some embodiments, the driver state assessment model can be implemented using a random forest. In this embodiment, the random forest improves robustness through voting or averaging of multiple decision trees, is relatively insensitive to outliers and noise, and is suitable for online scenarios with large fluctuations in traffic data.

[0218] In some embodiments, the driver state assessment model may be implemented using a support vector machine (SVM). In this embodiment, the SVM can find the maximum margin decision boundary based at least on the difference feature vectors, thereby achieving binary or multi-class classification of states.

[0219] In some embodiments, the driver state assessment model can be implemented using K-Nearest Neighbors (KNN). In this embodiment, the distance between the differential feature vector of the target vehicle and the differential features in the historical sample library is measured, and the K nearest samples are selected for voting or weighted voting to obtain the state result.

[0220] In some embodiments, the driver state assessment model may include a probabilistic statistical model. For explanation, a probabilistic statistical model can be used to calculate the posterior probability of the driver being in different states based on differential features and output a confidence level. In some embodiments, a Naive Bayes model may be employed. Specifically, differential features can be treated as conditional evidence, and the posterior probability of each state can be calculated based on the conditional independence assumption. In other embodiments, a Bayesian network can be constructed and a dependency structure can be established between differential features and between differential features and states, thereby enabling finer-grained probabilistic inference and uncertainty representation.

[0221] In some embodiments, the driver state assessment model may include a time-series model or a deep neural network model for processing sequences of differential features to output driver state assessment results using information about the evolution of differential features over time.

[0222] In some embodiments, when the input is a fixed-dimensional vector, a multilayer perceptron (MLP) can be used to learn a nonlinear mapping from differential features to state outcomes.

[0223] In some embodiments, when the input is a sequence of differential features, a recurrent neural network (RNN), a long short-term memory network (LSTM), or a gated recurrent unit (GRU) can be used to capture the dynamic process of risk accumulation or mitigation over time.

[0224] In some embodiments, attention mechanisms or Transformers can be used to assign higher weights to key windows in the differential feature sequences, thereby improving the ability to model long sequence dependencies and outputting more stable probabilities and confidence levels.

[0225] In some embodiments, while inputting the differential features into the driver state assessment model, other auxiliary variables may be further output to the driver state assessment model.

[0226] In some embodiments, the input to the driver state assessment model may also include one or more of the following: vehicle type information of the target vehicle, road segment feature information of the spatiotemporal trajectory, traffic flow coordination feature, and contextual feature.

[0227] In some embodiments, the input to the driver state assessment model may further include traffic flow coordination features. In this embodiment, traffic flow coordination features can be used to characterize the degree of consistency in the driving states of various vehicles in the current traffic flow. As an explanation, the driver state assessment model can comprehensively consider the numerical values ​​of the difference features and the confidence level reflected by the traffic flow coordination features to provide a more accurate assessment result.

[0228] In some embodiments, the traffic flow coordination characteristic can be determined based on the degree of dispersion of the baseline driving characteristics of the equivalent reference vehicle group, and is negatively correlated with the degree of dispersion. As an explanation, and not a limitation, when the speeds of all vehicles within the equivalent reference vehicle group are highly consistent (low dispersion), it indicates that the current traffic flow is in a coordinated state, with good driving order maintained among the vehicles; conversely, when the speeds of all vehicles within the equivalent reference vehicle group differ significantly (high dispersion), it indicates that the current traffic flow is in a non-coordinated state, and congestion, accidents, or other disruptive factors may exist.

[0229] In some embodiments, traffic flow coordination characteristics may include a dynamic weighting factor for traffic flow coordination. In some embodiments, the dynamic weighting factor for traffic flow coordination can be determined by the following formula:

[0230]

[0231] in, As a preset constant, The degree of dispersion of the baseline driving characteristics, This is the smoothing constant.

[0232] In some embodiments, a preset constant is used. This may include a constant used to characterize the standard deviation of speeds under standard road conditions. In this embodiment, a smoothing constant... It can be used to avoid cases where the denominator is zero. Its value can be set according to actual needs, such as 0.1 or 1.

[0233] In some embodiments, the dispersion of the baseline driving characteristics can be measured by the standard deviation of the baseline speed of an equivalent reference vehicle group. Characterization. As an explanation, and not a limitation, of the speed standard deviation in a comparable reference vehicle group. When the value is small, it indicates a high degree of coordination in the current traffic flow, and the reliability of the difference features is relatively high. Even slight deviations of the target vehicle may indicate an abnormal state, warranting greater attention. The speed standard deviation of the equivalent reference vehicle group... When the value is large, it indicates that the current traffic flow itself is in a chaotic state, and the confidence of the difference feature is reduced accordingly. At this time, the speed deviation of the target vehicle may be a normal response behavior, and it can be given higher attention. Therefore, the traffic flow coordination feature can be used as a reference factor for assessing the confidence of the difference feature.

[0234] In other embodiments, a dynamic weighting factor for traffic flow coordination can also be used to weight differential features instead of inputting them as independent variables into the driver state assessment model. In one example, the speed deviation can be multiplied by the dynamic weighting factor for traffic flow coordination to obtain a weighted speed deviation. By way of explanation and not limitation, this means that in scenarios with highly coordinated traffic flow, the speed deviation of the target vehicle will be amplified; in scenarios with chaotic traffic flow, the speed deviation of the target vehicle will be appropriately weakened. This dynamic weighting mechanism allows driver state assessment to better adapt to different traffic environments.

[0235] In some embodiments, the input to the driver state assessment model may further include situational features. In this embodiment, the situational features can be used to characterize the external prior risk situation in which the target vehicle is located at the current assessment moment.

[0236] In some embodiments, the contextual features may include at least one of the following: a time dimension feature determined based on the travel time of the target vehicle, a spatial dimension feature obtained based on the location information of the collection points passed by the target vehicle, and a continuous travel duration obtained based on the first travel record of the target vehicle.

[0237] In some embodiments, the time dimension feature may include a fatigue peak period identifier. In some embodiments, one or more fatigue peak periods may be preset. In a specific example, the fatigue peak periods are 0:00 to 5:00 AM and 1:00 to 3:00 PM. However, it is understood that the present invention does not limit the specific time period of the fatigue peak period. In other embodiments, those skilled in the art can select suitable fatigue peak periods based on historical accident statistics, road network operation experience, or time periods recommended by regulatory authorities, which falls within the protection scope of the present invention. In some embodiments, if the travel time recorded by the target vehicle currently falls within a preset fatigue peak period, the fatigue peak period identifier may be determined as a first value (e.g., 1); otherwise, it may be a second value (e.g., 0).

[0238] In some embodiments, spatial dimension features may include road segment attribute identifiers. In some embodiments i, the road segment attributes of the road segments traversed by the target vehicle can be determined by performing road network matching processing between the location information of the collection points and the geographic information system.

[0239] In some embodiments, the road segment attribute identifier may include a preset high-risk road segment identifier. In some embodiments, the road segment attribute identifier may include one or more of the following: an extra-long tunnel, a tunnel group, or a monotonous long straight road segment. However, it is understood that in other embodiments, the road segment attribute identifier may reasonably include more road segment attributes, and the present invention does not limit this. In some embodiments, if the road segment attribute identifier obtained by matching the target vehicle belongs to the preset high-risk road segment identifier, the road segment attribute identifier may be determined to be a first value (e.g., 1); otherwise, it may be a second value (e.g., 0).

[0240] In some embodiments, the cumulative continuous driving time can be used to characterize the continuous driving time of the target vehicle since the start of the current trip. In some embodiments, the current collection point and its corresponding timestamp, as well as the first collection point and its corresponding timestamp, of the target vehicle in the current trip can be determined, and the travel time between the current collection point and the first collection point can be determined as the cumulative continuous driving time. In other embodiments, if the time interval between two adjacent travel records exceeds a preset interval threshold, the current collection point and its corresponding timestamp, as well as the adjacent collection point and its corresponding timestamp, can be determined, and the travel time between the current collection point and the adjacent collection point can be determined as the cumulative continuous driving time.

[0241] In some embodiments, the driver condition assessment results may include a driver condition index and / or a risk level.

[0242] In some embodiments, the driver state index may include a continuous value. In one specific embodiment, the driver state index ranges from 0 to 100, where a higher value indicates a greater likelihood that the driver is in a fatigued or abnormal state.

[0243] In some embodiments, the risk level may include a discrete level determined based on a driver state index. In some embodiments, the risk level includes a normal state, a suspected fatigue state, and a confirmed fatigue state. In one specific embodiment, the risk level is determined according to the numerical range of the driver state index: a driver state index in the range of 0 to 60 corresponds to a normal state, a range of 60 to 80 corresponds to a suspected fatigue state, and a range of 80 to 100 corresponds to a confirmed fatigue state. However, it is understood that in other embodiments, the correspondence between the numerical range and the risk level can be adjusted as needed, and this invention does not limit this.

[0244] In some embodiments, the risk level threshold can be dynamically adjusted based on at least one of the following: the time period in which the target vehicle travels, the type of road segment traversed by the target vehicle, current weather conditions, etc. As an explanation and not a limitation, the high-risk threshold can be lowered during periods of high fatigue incidence (such as 0:00 to 5:00 AM), making it easier for the same behavioral characteristics to trigger a higher risk level during that period.

[0245] In some embodiments, when or after outputting the driver state assessment result corresponding to the target vehicle, the driver state assessment method may further include a step of intervening in the target vehicle and / or the target vehicle driver based on the driver state assessment result. Accordingly, refer to Figure 4 The driver status assessment method may further include the following step S410.

[0246] S410: In response to the driver status assessment result meeting the preset intervention conditions, perform intervention operations on the target vehicle and / or the driver of the target vehicle.

[0247] In some embodiments, the intervention includes measures taken to remind or alert the driver and prompt them to improve their driving behavior. In some embodiments, the intervention may be performed on a target vehicle (via in-vehicle equipment) and / or the driver (via information prompts).

[0248] In some embodiments, preset intervention conditions include conditions that trigger the intervention operation. In some embodiments, preset intervention conditions may include a driver's condition index exceeding a preset threshold, or a risk level reaching a preset level. In a specific example, an intervention operation is performed on the target vehicle and / or the driver of the target vehicle when the risk level is suspected fatigue or confirmed fatigue. In some embodiments, the intensity of the intervention operation may be positively correlated with the risk level.

[0249] In some embodiments, intervention operations may include one or more of the following: pushing prompt information and / or warning signals to the on-board terminal of the target vehicle, pushing targeted warning information to roadside information boards that the target vehicle is about to pass, and reporting high-risk vehicle information to the management center.

[0250] In some embodiments, the prompts and / or warning signals pushed to the in-vehicle terminal of the target vehicle may include voice prompts and / or a highlighted service area on the navigation interface. In one specific embodiment, the prompt message is "You have been driving continuously for a long time, it is recommended to go to a service area to rest." In some embodiments, the warning signals may include a rapid warning sound and / or seat vibration.

[0251] In some embodiments, the targeted warning information on the roadside information board may include: predicting the time when the target vehicle will arrive at the roadside information board based on the time the target vehicle passes the previous data collection point and the distance from the previous data collection point to the roadside information board; switching the display content of the roadside information board to the targeted warning information during a preset period before the predicted arrival time; and restoring the default display content of the information board after the target vehicle has passed. In a specific example, the targeted warning information on the roadside information board is "Driver of license plate number 123456, please pay attention to safety, there is a service area ahead for rest."

[0252] In some embodiments, information on high-risk vehicles can be reported to the management center, which can then arrange for manual tracking or dispatch road patrol forces based on the reported information.

[0253] In some embodiments, after performing an intervention operation on the target vehicle and / or the driver of the target vehicle, the effectiveness of the intervention operation can also be monitored to achieve intervention operation and closed-loop feedback. Accordingly, in some embodiments, after performing an intervention operation on the target vehicle and / or the driver of the target vehicle, the driver state assessment method may further include the following steps S420, S430, S440 and S450.

[0254] S420: Obtain the vehicle passage records of the target vehicle at subsequent collection points.

[0255] As previously described, the step of acquiring vehicle passage records collected by multiple vehicle-road cooperative communication devices deployed along the road can be performed in response to the time when a vehicle arrives at the current collection point. Accordingly, in some embodiments, subsequent collection points include the next (adjacent) collection point or multiple collection points after the current collection point and along the direction of travel of the target vehicle. For explanation, by acquiring the passage records of subsequent collection points, it is possible to understand the changes in the driving status of the target vehicle after intervention.

[0256] S430: Determine the subsequent driver status assessment results of the target vehicle based on subsequent vehicle passage records.

[0257] In some embodiments, the determination of subsequent driver status assessment results is similar to the aforementioned steps S110 to S160, except that the passage records used include passage records from subsequent collection points.

[0258] S440: Based on the subsequent driver status assessment results, evaluate the effectiveness of the intervention operation and obtain the intervention event including the effectiveness label.

[0259] In some embodiments, an intervention event may include a record of the intervention operation and related information. In some embodiments, an intervention event may include the intervention time, the intervention target, the intervention method, the driver's condition assessment results before the intervention, the driver's condition assessment results after the intervention, and an effectiveness label, etc.

[0260] In some embodiments, the effectiveness of the intervention can be assessed based on one or more of the examples above.

[0261] In some embodiments, the effectiveness label can be determined when the improvement in the subsequent driver status assessment result relative to the driver status assessment result before the intervention exceeds a preset threshold. In a specific example, the effectiveness label is determined if the driver status index after the intervention decreases by more than 15 points compared to before the intervention.

[0262] In some embodiments, validity can be determined when the target vehicle enters the service area and stays for more than a preset rest period. In a specific example, the validity label is determined to be valid if the target vehicle stays in the service area for more than 20 minutes after the intervention.

[0263] In some embodiments, the validity label may be determined to be invalid if subsequent driver condition assessments do not show significant improvement or deterioration.

[0264] In some embodiments, if the validity tag is determined to be invalid, a higher level of intervention may be performed on the target vehicle. In some embodiments, a warning signal may be issued to the target vehicle / driver. In some embodiments, roadside information board warnings may be issued and the information may be reported to the management center.

[0265] S450: Driver status assessment model updated based on intervention events.

[0266] In some embodiments, intervention events and their effectiveness labels can be used as new labeled data for retraining or updating the driver state assessment model.

[0267] In some embodiments, the step of updating the driver state assessment model based on intervention events may include the following steps E1 and E2. E1: Add the intervention event as new labeled data to the labeled dataset; E2: Retrain the driver state assessment model using the updated labeled dataset. For illustrative purposes, and not as limiting, events in which the driver's state actually improves after intervention can be used as positive samples to enhance the model's accuracy in identifying fatigue states; events in which the driver's state does not improve after intervention and no subsequent accidents occur can be used as negative samples to reduce the model's false alarm rate.

[0268] In the above embodiments of the present invention, through the closed-loop feedback mechanism of step intervention operation, the driver state assessment system can continuously accumulate actual operating data, thereby continuously optimizing the accuracy of the assessment model and achieving adaptive performance improvement.

[0269] Based on the solutions of the above embodiments, the driver state assessment method, system, and vehicle-road cooperative system provided by the present invention have at least the following technical effects:

[0270] First, in this embodiment of the invention, driver status assessment can be performed based on vehicle passage records collected by vehicle-road cooperative communication devices (such as ETC gantry systems and RSU roadside units). Since the vehicle passage records only include vehicle identification, passage time, and collection point location information, and do not involve the driver's facial images, physiological signals, or other sensitive personal information, this embodiment of the invention does not require privacy-sensitive technologies such as facial recognition and physiological monitoring, and can achieve driver status assessment function without infringing on the driver's personal privacy.

[0271] Meanwhile, in this embodiment of the invention, since the data required for driver status assessment comes from roadside infrastructure, there is no need to obtain driver information through cameras, wearable devices, or physiological sensors. Because my country's highways have largely achieved full coverage of ETC gantry systems, this embodiment of the invention can directly utilize existing infrastructure resources without the need for large-scale construction of new detection equipment. It has the advantages of low cost and rapid deployment, facilitating rapid promotion and application in the existing highway network, and achieving large-scale, all-weather driver status assessment capabilities.

[0272] In this embodiment of the invention, vehicles meeting equivalent conditions are determined at least based on vehicle traffic records, thereby constructing a set of equivalent reference vehicles and determining the baseline driving characteristics of the equivalent reference vehicle set, thus realizing a dynamically adaptive driver state assessment benchmark. Here, since the baseline driving characteristics of this embodiment are derived from the statistical processing results of the driving characteristics of vehicles currently meeting equivalent conditions, rather than pre-set fixed values, the baseline can be updated in real time with the current traffic flow state. For explanation, when encountering rain or snow that causes an overall decrease in vehicle speed, the baseline driving characteristics decrease accordingly, avoiding misjudging normally decelerating vehicles as abnormal; when encountering road construction or traffic congestion, the baseline driving characteristics are also automatically adjusted to ensure that the assessment result reflects the degree of deviation of the target vehicle relative to the current traffic flow, rather than the degree of deviation relative to a fixed standard. Furthermore, on the one hand, compared to known schemes using fixed thresholds, the scheme of this embodiment reduces the false alarm rate and false negative rate caused by changes in road conditions; on the other hand, compared to known schemes using historical average data as the baseline, the baseline of this embodiment can reflect the current real-time traffic state rather than outdated historical states, improving the timeliness and accuracy of the assessment.

[0273] In a further embodiment of the present invention, the combination of multi-dimensional equivalent conditions makes the equivalent reference vehicle group at least partially consistent with or similar to the target vehicle in terms of vehicle characteristics, road environment and time dimension. This eliminates the impact of differences in at least some factors among different vehicle models, such as power performance, road slope, curve curvature and speed limit regulations, on the baseline, thereby making the baseline driving characteristics statistically fair and providing a reliable reference benchmark for subsequent difference feature calculation.

[0274] In a further preferred embodiment of the present invention, by further filtering vehicles from vehicle traffic records that are similar to the target vehicle in terms of vehicle type, road segment, direction of travel, and passing through the same collection point within a preset time window, the construction of the equivalent reference vehicle group is ensured to the greatest extent possible to be reasonable and fair, thereby significantly improving the accuracy of subsequent driver status assessment.

[0275] In a further embodiment of the present invention, a peer reference vehicle group is constructed by means of vehicles passing through the same collection point within a preset time window. This peer reference vehicle group includes vehicles that pass through the same collection point within a preset time period before or after the target vehicle is detected, and is not limited to vehicles detected simultaneously with the target vehicle. Therefore, the construction of the peer reference vehicle group in this further embodiment does not depend on the presence of other vehicles around the target vehicle at the time of detection. This allows for the construction of a peer reference vehicle group with a sufficient sample size in scenarios with sparse traffic, such as late-night hours on highways, by accumulating vehicles passing through within the time window. Furthermore, compared to schemes that rely on a group of vehicles simultaneously detected by the same detection unit to construct a baseline, the solution of this embodiment overcomes the technical obstacle of not being able to construct an effective baseline when traffic is sparse, ensuring the reliability and continuity of the assessment, and achieving all-weather, all-road segment driver status assessment capabilities.

[0276] In a further embodiment of the present invention, the adaptive time window technique is used to dynamically adjust the time window according to the actual traffic flow, which further enhances the adaptability in sparse traffic scenarios. Thus, even during periods of extremely low traffic flow, it is possible to construct an equivalent reference vehicle group that meets the minimum sample size requirement, ensuring the statistical reliability of the baseline driving characteristics.

[0277] In an embodiment of the present invention, reference is made to Figure 5 The system also provides a driver state assessment system 500, which may include an acquisition module 510, a first determination module 520, a construction module 530, a second determination module 540, a third determination module 550, and an assessment module 560.

[0278] In some embodiments, the acquisition module 510 is configured to acquire vehicle passage records collected by multiple vehicle-road cooperative communication devices installed along the road, wherein the vehicle passage records include at least vehicle identification, passage time, and collection point location information. In some embodiments, the acquisition module 510 can connect to the vehicle-road cooperative communication devices or their data center via wired or wireless communication to acquire vehicle passage records. In some embodiments, the acquisition module 510 can receive passage records uploaded by the vehicle-road cooperative communication devices in real time. In optional embodiments, the acquisition module 510 can also periodically acquire passage records in batches from the data center.

[0279] In some embodiments, the first determining module 520 is configured to form a spatiotemporal trajectory or spatiotemporal trajectory segment of the target vehicle based on the vehicle passage records of the target vehicle, and determine the driving characteristics of the target vehicle in the spatiotemporal trajectory or spatiotemporal trajectory segment. In some embodiments, the first determining module 520 may filter the passage records belonging to the target vehicle from the vehicle passage records based on the vehicle identifier, sort and associate the filtered passage records according to the passage time, and form a spatiotemporal trajectory or spatiotemporal trajectory segment. In some embodiments, the first determining module 520 may also calculate the driving characteristics of the target vehicle, such as the interval average speed, based on the distance between adjacent collection points and the difference in passage time between the target vehicle and the adjacent collection points.

[0280] In some embodiments, the construction module 530 is configured to determine, at least based on vehicle traffic records, vehicles that meet preset equivalence conditions in order to construct a group of equivalence reference vehicles. In some embodiments, the construction module 530 may filter vehicles from vehicle traffic records that have the same or similar vehicle attributes as the target vehicle and / or are in the same or similar traffic environment conditions as the target vehicle, based on preset equivalence conditions.

[0281] In some embodiments, the second determining module 540 is configured to determine the baseline driving characteristics of a peer reference vehicle group. In some embodiments, the second determining module 540 may determine the driving characteristics of each vehicle in the peer reference vehicle group and perform statistical processing on the driving characteristics of each vehicle to obtain the baseline driving characteristics.

[0282] In some embodiments, the third determining module 550 is configured to determine the difference characteristics of the target vehicle relative to the equivalent reference vehicle group based on the driving characteristics of the target vehicle and the baseline driving characteristics of the equivalent reference vehicle group. In some embodiments, the third determining module 550 may calculate the degree of deviation of the driving characteristics of the target vehicle from the baseline driving characteristics to obtain the difference characteristics.

[0283] In some embodiments, the evaluation module 560 is configured to input at least the difference features into a preset driver state evaluation model to output the driver state evaluation result corresponding to the target vehicle. In some embodiments, the evaluation module 560 may input the difference features and optional other features (such as traffic flow coordination features, context features, etc.) into the preset driver state evaluation model, and the model outputs the driver state evaluation result.

[0284] In some embodiments, the driver state assessment system 500 may further include an intervention module (not shown in the figure), which is configured to perform an intervention operation on the target vehicle and / or the driver of the target vehicle in response to the driver state assessment result meeting a preset intervention condition.

[0285] In some embodiments, the driver state assessment system 500 may further include a feedback module (not shown in the figure), which is configured to acquire the vehicle passage records of the target vehicle at subsequent collection points, assess the effectiveness of the intervention operation based on the subsequent passage records, and update the driver state assessment model based on the intervention event.

[0286] In some embodiments, the driver state assessment system 500 may be deployed on a cloud server, an edge computing node, or a combination thereof. In some embodiments, the acquisition module 510 and the assessment module 560 may be deployed in the cloud for centralized data management and model operation; the construction module 530 and the second determination module 540 may be deployed on an edge computing node close to the acquisition point to reduce data transmission latency and bandwidth consumption.

[0287] In an embodiment of the present invention, reference is made to Figure 6 The system also provides a vehicle-road cooperative system 600, which may include a vehicle-road cooperative communication device 610, an on-board unit 630, and a driver status assessment system 500 as described in any of the foregoing embodiments.

[0288] In some embodiments, multiple vehicle-road cooperative communication devices 610 may be installed along a road, and the vehicle-road cooperative communication devices 610 are configured to collect vehicle passage records and send the vehicle passage records to a driver status assessment system 500. In some embodiments, the vehicle-road cooperative communication devices 610 may be distributed along the road direction, and there may be a certain distance between adjacent devices.

[0289] In some embodiments, the vehicle-to-infrastructure (V2I) communication device 610 may include a roadside unit (RSU) 620. In this embodiment, the RSU 620 can communicate with the on-board unit 630 via a V2I communication protocol such as DSRC or C-V2X to obtain relevant vehicle information when a vehicle passes by. In an optional embodiment, the V2I communication device 610 may further include an ETC gantry system, which can obtain vehicle passage records by interacting with the on-board ETC tag when a vehicle passes by.

[0290] In some embodiments, the vehicle passage records collected by the vehicle-road cooperative communication device 610 include at least vehicle identification, passage time, and collection point location information.

[0291] In some embodiments, the on-board unit 630 is installed in a vehicle and configured to communicate with the vehicle-to-infrastructure (V2I) communication device 610 and receive intervention information sent by the driver state assessment system 500. In some embodiments, the on-board unit 630 may be, for example, but not limited to, an on-board unit (OBU), an ETC tag, or other devices capable of communicating with the V2I communication device 610. In some embodiments, the on-board unit 630 may interact with the V2I communication device 610 when the vehicle passes by, enabling the V2I communication device 610 to obtain the vehicle's identification information and travel time.

[0292] In some embodiments, the vehicle unit 630 may also have information receiving and display functions. In some embodiments, the vehicle unit 630 may receive intervention information sent by the driver state assessment system 500 and convey it to the driver through voice broadcast, screen display, vibration prompts, etc.

[0293] In some embodiments, the vehicle-to-infrastructure (V2I) system 600 may further include roadside information display devices (not shown), such as variable message signs. In some embodiments, the roadside information display devices are configured to receive targeted warning information from the driver status assessment system 500 and display it to the driver of a specific vehicle.

[0294] In some embodiments, the vehicle-road cooperative system 600 may further include a management center (not shown in the figure). The management center is configured to receive high-risk vehicle information reported by the driver status assessment system 500 and to dispatch human intervention forces as needed.

[0295] The vehicle-road cooperative system 600 provided in the above embodiments of the present invention can make full use of existing vehicle-road cooperative infrastructure and achieve non-intrusive assessment and intervention of driver status without additional vehicle modifications. Simultaneously, the vehicle-road cooperative system 600, through peer reference vehicle groups and a dynamic baseline mechanism, can accurately assess driver status in various traffic environments and effectively improve road traffic safety by transmitting intervention information through onboard units and roadside information dissemination equipment via multiple channels via onboard units and roadside information dissemination devices.

[0296] The driver state assessment system, vehicle-road cooperative system, and their components, modules, units, and features described in the embodiments of the present invention can be combined with the driver state assessment method of the embodiments of the present invention in a non-contradictory manner to obtain new embodiments, which will not be elaborated here. Conversely, the driver state assessment method, its steps, sub-steps, and features described in the embodiments of the present invention can also be combined with the driver state assessment system and vehicle-road cooperative system of the embodiments of the present invention in a non-contradictory manner to obtain new embodiments.

[0297] In some embodiments of the present invention, an electronic device is provided, including a processor and a memory storing a computer program, the processor being configured to implement the method of any embodiment of the present invention when running the computer program.

[0298] Figure 7 A schematic diagram is shown of a method or electronic device 700 that can be used to implement embodiments of the present invention. In some embodiments, the number of electronic devices may be more or less than the number shown. In some embodiments, a single or multiple electronic devices may be used. Cloud-based or distributed electronic devices may also be used in some embodiments.

[0299] like Figure 7 As shown, the computer 700 includes a processor 701, which can perform various appropriate operations and processes based on programs and / or data stored in read-only memory (ROM) 702 or programs and / or data loaded from storage portion 708 into random access memory (RAM) 703. The processor 701 may include a central processing unit (CPU), and may be a multi-core processor or contain multiple processors. In some embodiments, the processor 701 may include a general-purpose main processor and one or more special coprocessors, such as a graphics processing unit (GPU), a neural network processor (NPU), a digital signal processor (DSP), etc. Various programs and data required for the operation of the electronic device 700 are also stored in RAM 703. The processor 701, ROM 702, and RAM 703 are interconnected via a bus 704. An input / output (I / O) interface 705 is also connected to the bus 704.

[0300] The processor and memory described above are used together to execute a program stored in the memory. When the program is executed by a computer, it can implement the steps or functions of the methods described in the above embodiments.

[0301] The following components are connected to the I / O interface 705: an input section 706 including a keyboard, mouse, touchscreen, etc.; an output section 707 including a cathode ray tube (CRT), liquid crystal display (LCD), etc., and speakers, etc.; a storage section 708 including a hard disk, etc.; and a communication section 709 including a network interface card such as a LAN card, modem, etc. The communication section 709 performs communication processing via a network such as the Internet. A drive 710 is also connected to the I / O interface 705 as needed. A removable medium 711, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., is installed on the drive 710 as needed so that computer programs read from it can be installed into the storage section 708 as needed. Figure 7 The diagram only shows a portion of the components and does not imply that the computer system 700 includes only a portion of the components. Figure 7 The components shown.

[0302] Although not shown, in this embodiment of the invention, a program product is provided, including a computer program that, when executed by a processor, implements the method of any embodiment of the invention.

[0303] Although not shown, in an embodiment of the invention, a storage medium is provided storing a computer program configured to be run to implement the method of any embodiment of the invention.

[0304] Storage media in embodiments of the present invention include articles that are permanent and non-permanent, removable and non-removable, capable of storing information by any method or technology. Examples of 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 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.

[0305] The methods, programs, systems, apparatuses, etc., in embodiments of the present invention can be executed or implemented in one or more networked computers, or practiced in a distributed computing environment. In the embodiments of this specification, in these distributed computing environments, tasks can be performed by remote processing devices connected via a communication network.

[0306] Those skilled in the art will understand that the embodiments described in this specification can be provided as methods, systems, or computer program products. Therefore, those skilled in the art will realize that the functional modules / units or controllers and related method steps described in the above embodiments can be implemented in software, hardware, or a combination of both.

[0307] Unless explicitly stated otherwise, the actions or steps of the methods and procedures described in the embodiments of the present invention do not necessarily have to be performed in a specific order and can still achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0308] This document describes several embodiments of the present invention; however, for the sake of brevity, the descriptions of the embodiments are not exhaustive, and identical or similar features or parts between the embodiments may be omitted. In this document, "one embodiment," "some embodiments," "example," "specific example," or "some examples" refers to embodiments applicable to at least one, but not all, of the present invention. The above terms do not necessarily refer to the same embodiments or examples. Without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described herein, as well as the features of the different embodiments or examples.

[0309] The exemplary systems and methods of the present invention have been specifically shown and described with reference to the above embodiments, which are merely examples of the best mode for implementing the systems and methods. Those skilled in the art will understand that various changes can be made to the embodiments of the systems and methods described herein without departing from the spirit and scope of the invention as defined in the appended claims when implementing the systems and / or methods.

Claims

1. A method for assessing driver condition, characterized in that, The method includes: The vehicle passage records are collected by multiple vehicle-road cooperative communication devices set up along the road, wherein the vehicle passage records include at least vehicle identification, passage time and collection point location information; Based on the vehicle passage records of the target vehicle, a spatiotemporal trajectory or spatiotemporal trajectory segment of the target vehicle is formed, and the driving characteristics of the target vehicle in the spatiotemporal trajectory or spatiotemporal trajectory segment are determined. At least vehicles that meet the preset equivalence conditions are identified based on the vehicle passage records to construct an equivalence reference vehicle group; Determine the baseline driving characteristics of the equivalent reference vehicle group; Based on the driving characteristics of the target vehicle and the baseline driving characteristics of the equivalent reference vehicle group, the difference characteristics of the target vehicle relative to the equivalent reference vehicle group are determined. The difference features are at least input into a preset driver state assessment model to output the driver state assessment result corresponding to the target vehicle.

2. The method according to claim 1, characterized in that, Determining the driving characteristics of the target vehicle in the spatiotemporal trajectory or spatiotemporal trajectory segment includes: The first driving characteristic of the target vehicle is calculated based on the distance between at least two collection points and the travel time of the target vehicle through the at least two collection points; The determination of the baseline driving characteristics of the equivalent reference vehicle group includes: The first reference driving characteristics of each peer reference vehicle are calculated based on the distance between the at least two collection points and the travel time of each peer reference vehicle between the at least two collection points. The first baseline driving feature is obtained by statistically processing the first reference driving feature.

3. The method according to claim 1, characterized in that, Determining the driving characteristics of the target vehicle in the spatiotemporal trajectory or spatiotemporal trajectory segment includes: Calculate multiple first driving characteristics of the target vehicle based on the distance between at least three collection points and the travel time of the target vehicle through the at least three collection points; Statistical processing is performed on the multiple first driving characteristics to obtain statistical driving characteristics; Based on the plurality of first driving characteristics and the statistical driving characteristics, the second driving characteristics of the target vehicle are determined; The determination of the baseline driving characteristics of the equivalent reference vehicle group includes: Based on the distance between the at least three collection points and the travel time of each peer reference vehicle through the at least three collection points, multiple first reference driving characteristics of each peer reference vehicle are calculated. The statistical reference driving characteristics of each peer reference vehicle are obtained by statistically processing the multiple first reference driving characteristics of each peer reference vehicle. Based on the plurality of first reference driving characteristics and the statistical reference driving characteristics, the second reference driving characteristics of each equivalent reference vehicle are determined; The second baseline driving characteristics are obtained by statistically processing the second reference driving characteristics.

4. The method according to claim 2 or 3, characterized in that, The first driving characteristic includes vehicle speed; The difference features include vehicle speed deviation, which characterizes the relative deviation between the target vehicle's speed and the average speed of the equivalent reference vehicle group, and is determined by the following formula: in, For vehicle speed deviation, The average speed of the target vehicle between two adjacent data collection points. The baseline average speed of the equivalent reference vehicle group on the same road segment.

5. The method according to claim 3, characterized in that, The second driving characteristic includes driving smoothness, which characterizes the rate of change of vehicle speed across multiple consecutive road segments, and the driving smoothness of the target vehicle is determined by the following formula: in, For the driving smoothness of the target vehicle. Let be the average speed of the target vehicle in the j-th target road segment. For the target vehicle in Average vehicle speed on each road segment Number of road segments; The driving smoothness of each equivalent reference vehicle is determined by the following formula: in, For the driving smoothness of each equivalent reference vehicle, Let be the interval average speed of each equivalent reference vehicle on the i-th target road segment. For each equivalent reference vehicle in Average vehicle speed on each road segment Number of road segments; The differential characteristics include a smoothness anomaly index, which is determined by the following formula: in, The smoothness anomaly index for the target vehicle. For the driving smoothness of the target vehicle. For the driving smoothness of each equivalent reference vehicle The average value.

6. The method according to claim 1, characterized in that, The method further includes: Traffic flow coordination characteristics are determined based on the degree of dispersion of the baseline driving characteristics of the peer reference vehicle group, wherein the traffic flow coordination characteristics are negatively correlated with the degree of dispersion; The step of inputting the difference features into a preset driver state assessment model to output the driver state assessment result corresponding to the target vehicle includes: inputting the difference features and the traffic flow coordination feature into a preset driver state assessment model to output the driver state assessment result corresponding to the target vehicle.

7. The method according to claim 1, characterized in that, The method further includes: In response to the driver status assessment result meeting the preset intervention conditions, an intervention operation is performed on the target vehicle and / or the driver of the target vehicle; Obtain the vehicle passage records of the target vehicle at subsequent collection points, and determine the subsequent driver status assessment result of the target vehicle based on the subsequent vehicle passage records; Based on the subsequent driver status assessment results, the effectiveness of the intervention is evaluated to obtain intervention events including effectiveness labels; The driver status assessment model is updated based on the intervention event.

8. The method according to claim 1, characterized in that, The process of determining vehicles that meet preset equivalence conditions to construct an equivalent reference vehicle group includes: The vehicle reference vehicle group is constructed by filtering vehicles from the vehicle traffic records that are the same or similar to the target vehicle in terms of vehicle type, road segment, and direction of travel, and that pass through the same collection point within a preset time window.

9. A driver condition assessment system, characterized in that, include: The acquisition module is configured to acquire vehicle passage records collected by multiple vehicle-road cooperative communication devices set up along the road, wherein the passage records include at least vehicle identification, passage time and collection point location information; The first determining module is configured to form the spatiotemporal trajectory or spatiotemporal trajectory segment of the target vehicle based on the vehicle passage records of the target vehicle, and determine the driving characteristics of the target vehicle in the spatiotemporal trajectory or spatiotemporal trajectory segment. The module is configured to determine at least the vehicles that meet the preset equivalence conditions based on the vehicle passage records, so as to construct an equivalence reference vehicle group; The second determining module is configured to determine the baseline driving characteristics of the peer reference vehicle group; The third determining module is configured to determine the difference characteristics of the target vehicle relative to the equivalent reference vehicle group based on the driving characteristics of the target vehicle and the baseline driving characteristics of the equivalent reference vehicle group. The evaluation module is configured to input the difference features into a preset driver state evaluation model to output the driver state evaluation result corresponding to the target vehicle.

10. A vehicle-road cooperative system, characterized in that, include: Multiple vehicle-road cooperative communication devices are installed along the road, and the vehicle-road cooperative communication devices include RSU roadside units; Onboard unit (OBU) installed in the vehicle; as well as The driver status assessment system according to claim 9.

Images