Traffic assistance method and device based on driver traffic accident risk prediction

Abstract

The application provides a traffic assistance method and device based on driver traffic accident risk prediction, which is applied to a traffic assistance system, the traffic assistance system comprises a vehicle terminal, a roadside device and a server, and the method comprises the following steps: the roadside device receives driving information of each vehicle; the roadside device obtains driver comprehensive scores of each driver from the server based on the driver information of each vehicle; for each group of adjacent vehicles, the roadside device predicts whether each group of the adjacent vehicles has a collision risk based on driving data and the driver comprehensive scores of the adjacent vehicles; a collision risk event is generated for each vehicle in the adjacent vehicles having the collision risk; and the roadside device sends the collision risk event to the adjacent vehicles having the collision risk. Based on this, the roadside device can perform multi-source collaborative early warning in combination with data of a cloud server and data of on-site vehicles, the prediction accuracy is improved, the accident reaction window is prolonged, and the accident rate is reduced.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of network communication, in particular to a traffic assistance method and device based on driver traffic accident risk prediction. BACKGROUND

[0002] At present, driving has become the main way of people's travel. But in the process of driving, different driving habits will have different driving risks. And the emotions of the driver at different times will also affect the driving behavior and bring unnecessary risks.

[0003] In order to improve the driving safety at present, an automatic emergency braking system (Autonomous Emergency Braking, AEB) is generally installed on the vehicle, which triggers AEB to stop in emergency, which improves the survival rate of the driver and passenger to a certain extent.

[0004] However, the risk prediction of this way is usually some emergency, sometimes only reduces the damage of the accident, and cannot avoid the occurrence of the accident. How to reduce the probability of the occurrence of the accident is a technical problem which the person skilled in the art urgently needs to solve. SUMMARY

[0005] Therefore, the purpose of the embodiments of the present application is to provide a traffic assistance method and device based on driver traffic accident risk prediction to solve the problems existing in the prior art.

[0006] In a first aspect, an embodiment of the present application provides a traffic assistance method based on driver traffic accident risk prediction, the method being applied to a traffic assistance system, the traffic assistance system comprising a vehicle terminal, a roadside device and a server, the method comprising: the roadside device receiving driving information of each vehicle, the driving information comprising driver information and driving data; wherein the vehicle terminal acquires the driving information and sends the driving information to the roadside device, and the roadside device reports to the server; the roadside device acquires driver comprehensive scores of each driver from the server based on the driver information of each vehicle, the driver comprehensive score being determined based on a benchmark score and a personnel score, the benchmark score being determined based on an accident involvement rate, an accident frequency rate and an accident intensity rate, the accident involvement rate being a proportion of drivers involved in at least one accident in a period to total drivers, the accident frequency rate being an accident rate per person, and the accident intensity rate being a repeated number of people involved in accidents, and the personnel score being used to evaluate a personal driving risk score of a driver; the driving data comprising space-time information and environmental data; for each group of adjacent vehicles, the roadside device predicts whether each group of adjacent vehicles has a collision risk based on the driving data and the driver comprehensive score of the adjacent vehicles; for each vehicle in the adjacent vehicles having a collision risk, a collision risk event is generated respectively, the collision risk event comprising a position of an opposite vehicle having a collision risk, a predicted behavior of the opposite vehicle causing a collision and a predicted behavior of the self causing a collision; and the roadside device sends the collision risk event to the adjacent vehicles having a collision risk.

[0007] In a second aspect, a traffic assistance device based on driver traffic accident risk prediction is provided. The device is applied to a traffic assistance system, which comprises a vehicle terminal, a roadside device and a server. The device comprises: a receiving module configured to receive driving information of each vehicle, wherein the driving information comprises driver information and driving data; wherein the vehicle terminal acquires the driving information and sends the driving information to the roadside device, and the roadside device reports the driving information to the server; an obtaining module configured to obtain driver comprehensive scores of each driver from the server based on the driver information of each vehicle, wherein the driver comprehensive score is determined based on a benchmark score and a personnel score, the benchmark score is determined based on an accident involvement rate, an accident frequency rate and an accident intensity rate, the accident involvement rate is a proportion of drivers involved in at least one accident in a period to total drivers, the accident frequency rate is an accident rate per person, and the accident intensity rate is a repeated number of people involved in accidents, and the personnel score is used to evaluate a personal driving risk score of a driver; the driving data comprises space-time information and environmental data; a prediction module configured to, for each group of adjacent vehicles, predict whether each group of the adjacent vehicles has a collision risk based on the driving data and the driver comprehensive score of the adjacent vehicles; generate a collision risk event for each vehicle in the adjacent vehicles having the collision risk, wherein the collision risk event comprises a position of an opposite vehicle having the collision risk, a predicted behavior of the opposite vehicle causing a collision and a predicted behavior of a self vehicle causing the collision; and a sending module configured to send the collision risk event to the adjacent vehicles having the collision risk.

[0008] In the method and apparatus provided in this embodiment of the invention, the roadside equipment receives driving information from each vehicle, the driving information including driver information and driving data; wherein, the vehicle-mounted terminal acquires the driving information and sends the driving information to the roadside equipment, which then reports it to the server; the roadside equipment obtains a comprehensive driver score for each driver from the server based on the driver information of each vehicle, the comprehensive driver score being determined based on a baseline score and a personnel score, the baseline score being determined based on accident involvement rate, accident frequency rate, and accident intensity rate, the accident involvement rate being the proportion of drivers involved in at least one accident within a period out of the total number of drivers, and the accident frequency rate being the proportion of drivers involved in at least one accident within a period out of the total number of drivers. The average accident rate and accident intensity rate are the number of times an accident occurs, respectively. The driver's score is used to evaluate their individual driving risk score. The driving data includes spatiotemporal information and environmental data. For each group of adjacent vehicles, the roadside equipment predicts whether there is a collision risk based on the driving data of the adjacent vehicles and the driver's comprehensive score. For each of the adjacent vehicles with a collision risk, a collision risk event is generated. This collision risk event includes the position of the other vehicle with a collision risk, the predicted collision-causing behavior of the other vehicle, and the predicted collision-causing behavior of the other vehicle. The roadside equipment sends the collision risk event to the adjacent vehicles with a collision risk. Based on this, the roadside equipment can combine data from the cloud server and on-site vehicle data for multi-source collaborative early warning, improving prediction accuracy, extending the accident response window, and reducing the accident rate.

[0009] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0010] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0011] Figure 1 A flowchart of a traffic assistance method based on driver traffic accident risk prediction provided by an embodiment of the present invention is shown;

[0012] Figure 2 A schematic diagram of the structure of a traffic assistance device based on driver traffic accident risk prediction provided in an embodiment of the present invention is shown;

[0013] Figure 3 A schematic diagram of the structure of an electronic device provided in an embodiment of the present invention is shown. Detailed Implementation

[0014] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0015] Figure 1 This is a schematic flowchart of a traffic assistance method based on driver traffic accident risk prediction, provided as an embodiment of the present invention. The method is applied to a traffic assistance system, which includes an onboard terminal, roadside equipment, and a server. The method is applied to the roadside equipment, which can monitor and communicate with all vehicles equipped with onboard terminals within its jurisdiction.

[0016] A traffic accident is an unexpected event that occurs on a road involving vehicles or pedestrians, which may result in personal injury or property damage. Traffic accidents are often closely related to driving behavior, environmental factors, and other factors.

[0017] Roadside equipment is a system that uses technological means (such as cameras and sensors) to observe and record specific areas in real time. It can also be used to receive driving information from vehicle terminals via wireless communication. In this paper, roadside equipment is used to monitor vehicles and predict potential traffic accident risks.

[0018] Jurisdiction refers to the area under the responsibility of a particular management agency or system. In this context, jurisdiction specifically refers to the road area covered by roadside equipment signals.

[0019] like Figure 1 As shown, the method includes:

[0020] S110, roadside equipment receives driving information from each vehicle.

[0021] The driving information includes driver information and driving data; the vehicle terminal can acquire the driving information and send it to the roadside equipment, which then reports it to the server for aggregation and statistics.

[0022] S120, the roadside equipment obtains the comprehensive driver score of each driver from the server based on the driver information of each vehicle.

[0023] The driver's comprehensive score can be determined based on a baseline score and personnel scores. The baseline score can be determined based on the accident involvement rate, accident frequency rate, and accident intensity rate. The accident involvement rate is the proportion of drivers involved in at least one accident within a period of time. The accident frequency rate is the average accident rate per person. The accident intensity rate is the number of times a person is involved in an accident.

[0024] This score is used to evaluate a driver's individual driving risk score.

[0025] The driving data can include spatiotemporal information as well as environmental data.

[0026] The driver's overall score is a comprehensive score used to measure the overall risk level of a driver within a specific time period, reflecting the likelihood that the driver may cause a traffic accident.

[0027] The baseline score is the foundation of a driver's overall rating. Its purpose is to provide an objective risk assessment standard and help compare the risk levels of different drivers.

[0028] Accident involvement rate can reflect the overall safety status of the driver population in a certain area.

[0029] The driver score is a rating system designed for individual drivers to assess their personal driving risk. It may incorporate personalized data such as driving habits and violation records to supplement the baseline score.

[0030] Current driving data includes spatiotemporal information and environmental data.

[0031] Spatiotemporal information refers to data on the time and location of driving behavior, as well as other behavioral data. This information can help analyze changes in a driver's risk over different time periods or areas, such as performance during peak hours or in complex road conditions.

[0032] In some embodiments, the server periodically collects driving data from roadside equipment for each driver, and determines a baseline score and a personal score based on the driving data of each driver. Determined based on the following formula:

[0033] ;

[0034] in, For accident involvement rate, For accident frequency, For accident intensity rate, For fluctuation penalty coefficient, This is the strength coefficient.

[0035] The driver's overall score is determined based on the following formula:

[0036] Score=factor coefficient WOE value

[0037] wherein, factor is based on the baseline score conversion determination, the total WOE value is equal to the sum of the values of the individual driving for all key bins; WOE for each key bin = ln (proportion of high-risk drivers in the bin / proportion of general-risk drivers in the bin); each key bin corresponds to a key feature.

[0038] The key features include one or more of the following: gender, whether holding A license, age group, driving experience group, cumulative score, number of accidents in the past 3 years, and number of violations in the past 3 years.

[0039] The key features also include one or more of the following: frequency of sudden braking, proportion of lane deviation duration, frequency of time distance less than 2s, accident rate correction coefficient at night, visual deviation duration, physiological fatigue degree, and violation type.

[0040] As an example, a structured feature reflecting the driving behavior characteristics and historical risk of the driver can be built. 21-23 years of violation records and accident experiences of 1.8 million drivers can be randomly sampled. Invalid fields (such as the number of violations for other road reasons, etc.) are excluded. Then, binning processing is performed, for example:

[0041] Age segmentation: '0_20 years old and below', '1_21-25 years old', '2_26-35 years old', '3_36-45 years old', '4_46-55 years old', '5_56-70 years old', '6_70 years old and above'.

[0042] Driving experience layer: '0_new driver (0-2 years)', '1_beginner (3-5 years)', '2_intermediate (6-10 years)', '3_senior (11-20 years)', '4_old driver (more than 20 years).

[0043] Violation penalty number binning: use decision tree algorithm to find the optimal cutting point.

[0044] Feature encoding: convert continuous variables into categorical labels (such as the number of speeding violations → 0 times / 1-3 times / 4+ times).

[0045] Output results: pre-processed feature matrix (including target variable labels).

[0046] Then, quantify the contribution of different feature values to the risk of accidents. The weight of evidence (WOE) of each bin can be calculated: WOE = ln (proportion of high-risk drivers in the bin / proportion of general-risk drivers in the bin). Then, keep the features with information value > 0.02 (such as fatigue driving time, number of night accidents) as key features.

[0047] Where, positive WOE value generally means risk increases; negative WOE value generally means risk decreases.

[0048] As an example, the output result: feature WOE mapping table is shown in Table 1;

[0049] Table 1

[0050]

[0051] Where, for The calculation formula is mainly through logarithmic transformation to reduce the skewness of frequency rate (FR), multiply the involvement rate (IR) to enhance the linkage of indicators, and combine the fluctuation penalty term to get more accurate values.

[0052] In some embodiments, the initial benchmark score is 600 points (corresponding to a benchmark risk ratio of 50:1); the score scale is: increase 20 points for every 2-fold reduction in risk; feature score: calculate the score of each bin by WOE and model coefficient.

[0053] Based on this, the score card configuration and calculation process is shown in Table 2:

[0054] Table 2

[0055]

[0056] Person score (Score) WOE value) example, assuming a person is shown in Table 3:

[0057] Table 3

[0058]

[0059] Then the score of this person is: Score = final benchmark score (773.8) + ∑ (bin variable score) = 773.8 + (-2.6369) + (0.9924) + (-10.0417) + (-9.3043) + (0.8633) + (5.4204) + (7.6092) = 773.8 + (-7.0976) = 766.7024.

[0060] S130, for each group of adjacent vehicles, the roadside device predicts whether each group of adjacent vehicles has a collision risk based on the driving data and the driver comprehensive score of the adjacent vehicles; and a collision risk event is generated for each vehicle in the adjacent vehicles that has a collision risk.

[0061] The collision risk event can include the position of the opposite vehicle that has a collision risk, the predicted behavior of the opposite vehicle causing a collision, and the predicted behavior of the self causing a collision.

[0062] The interaction risk index and the scene risk index of each group of adjacent vehicles can be predicted based on the driving data and the driver comprehensive score of the adjacent vehicles for each group of adjacent vehicles; and whether each group of adjacent vehicles has a collision risk is predicted based on the interaction risk index and the scene risk index of each group of adjacent vehicles.

[0063] In some embodiments, the interaction risk index is determined based on the following formula: The interaction risk index is determined based on the following formula:

[0064]

[0065] wherein, is the driver comprehensive score of the current vehicle; is the driver comprehensive score of the adjacent vehicle; is the time coupling degree , is the time difference to the conflict point, is the reference time; is the space coupling degree , is the vehicle distance, is the reference distance; is the scene amplification coefficient.

[0066] The scene risk index is determined based on the following formula:

[0067]

[0068] wherein, is the weather basic risk, obtained based on meteorological application; is the visibility parameter, obtained based on data analysis of vehicle radar point cloud; is the road alignment risk, determined based on map information; is the visual obstruction coefficient, determined based on the image obtained by the vehicle-mounted camera.

[0069] ​​​In some embodiments, the product of the interaction risk index and the scenario risk index can be calculated, and a comparison is made between the product and a pre-set threshold value. When the threshold value is exceeded, it is determined that there is a risk of collision. The product can be a numerical value that quantitatively assesses the risk that the vehicle can face in the current driving situation, taking into account various factors such as driver behavior, vehicle interaction, and external environment. The higher the score, the greater the potential risk in the current driving environment.

[0070] The interaction risk index (IRE) is used to measure the potential risk between the current vehicle and the adjacent vehicle. It is calculated based on the risk comprehensive index of both drivers, the time coupling degree, and the space coupling degree. The time coupling degree reflects the time difference of the two vehicles reaching the conflict point, and the space coupling degree represents the distance relationship between the two vehicles.

[0071] The scenario risk index mainly reflects the influence of the external environment on driving safety.

[0072] The scenario amplification coefficient is a parameter used to adjust the interaction risk index, which can further amplify or reduce the risk according to the characteristics of different scenarios. For example, in a complex traffic environment, the value of λ can be higher to reflect a higher risk level.

[0073] The behavior of the other party causing a collision or the behavior of the other party causing a collision includes one or more of the following: sudden braking, lane deviation, distance reduction, visual deviation, fatigue driving, and illegal behavior.

[0074] The specific collision behavior can be predicted according to the risk factor that contributes most to the interaction risk and the scenario risk.

[0075] For a risky vehicle pair (vehicle A and vehicle B), the roadside device will generate a collision risk event for them respectively.

[0076] Taking the perspective of vehicle A (the other party) as an example: the predicted behavior of the other party (vehicle B) causing a collision: the real-time data and high-risk features of vehicle B will be analyzed. If the driver's comprehensive score (Score B ) of vehicle B is very low, indicating that the driver has bad driving habits. The key features can be combined to make predictions:

[0077] If the "sudden braking frequency" in its historical data is high, it can be predicted that the other party has a "sudden braking" behavior.

[0078] If the "lane deviation duration proportion" is high, it can be predicted that the other party has a "lane deviation" behavior.

[0079] If the "fatigue driving" physiological indicators are abnormal, it can be predicted that the other party has a "fatigue driving" behavior.

[0080] Based on real-time driving data from vehicle B:

[0081] If the radar detects that vehicle B is rapidly approaching vehicle A, it predicts that the other vehicle will cause the "distance to decrease".

[0082] If the camera detects that vehicle B's trajectory is crossing the line, it predicts that the other vehicle is "deviating from its lane".

[0083] Based on scenario risks:

[0084] In low-visibility conditions, it can predict that the other party may "suddenly brake" or "drift out of lane" due to poor visibility.

[0085] Predicted behavior of our vehicle (Vehicle A) that would cause a collision:

[0086] The analysis can focus on vehicle A's real-time data and high-risk characteristics, as well as the responsibility of the vehicle in the interaction.

[0087] If the driver's overall score for vehicle A is... A If the risk level is too low, analyze A's high-risk characteristics to predict its possible behavior.

[0088] Based on real-time driving data from vehicle A:

[0089] If the system detects that vehicle A is following too closely, it will predict that the following distance of its own vehicle will decrease.

[0090] If the system detects that the driver of vehicle A has a deviated gaze (through the in-vehicle camera, the recognition result is directly collected and identified through the vehicle terminal), it will predict that the driver of vehicle A has a "deviated gaze".

[0091] If vehicle A is behind another vehicle and traveling at a relatively high speed during the interaction, it may be predicted that the other vehicle will not be able to brake in time.

[0092] Based on interaction relationships:

[0093] In merging lanes or intersections, if the arrival time of one vehicle at the point of conflict is very close to that of the other vehicle, the system may predict that the vehicle has committed a "violation" (such as failing to yield as required).

[0094] S140, the roadside equipment sends collision risk events to adjacent vehicles that are at risk of collision.

[0095] The vehicle-mounted terminal receives collision risk events from roadside equipment and issues warnings via the vehicle's voice system. The warning information, which may include the collision risk event, can also be displayed on the central control screen.

[0096] The roadside device can combine the data of the cloud server and the data of the vehicle on the spot to perform multi-source cooperative early warning, improve the prediction accuracy, extend the accident reaction window, and reduce the accident rate.

[0097] For the benchmark score calculation, the multi-dimensional quantification of accidents is performed, the group risk distribution is reflected through the involvement rate (I), the accident occurrence rate is measured through the frequency rate (F), the accident severity is evaluated through the intensity rate (S), and the misjudgment caused by data fluctuation is inhibited through the fluctuation penalty mechanism, and the system stability is improved.

[0098] The key features are binned through the WOE binning technology to realize nonlinear relationship modeling. The traditional linear scoring card is sensitive to extreme values, and the binning rule is clear for the risk contribution of each feature.

[0099] The interaction risk model is quantified through space-time coupling, the time coupling degree solves the trajectory conflict delay prediction problem, the space coupling degree establishes a distance-risk attenuation model, and the risk resonance suppression is realized. The early warning accuracy of double high-risk vehicles in intersection scenes is improved.

[0100] For the scene risk calculation environment factor fusion, the precision is improved. Among them, the meteorological risk base value can be determined based on meteorological API + sensor fusion, the visibility dynamic correction can be determined based on laser radar point cloud analysis, the road alignment risk compensation can be determined based on high-precision map curvature integration, and the field of view obstruction enhancement coefficient can be determined based on visual semantic segmentation.

[0101] In some embodiments, the binning rule of each key feature is:

[0102] The binning boundary is {[0, m-n]: low-risk bin; (n-m, n]: low-medium-risk bin; (n, n+m]: medium-high-risk bin; (n+m, +∞): high-risk bin}; wherein n is the historical mean value of the feature, and m is the standard deviation.

[0103] In some embodiments, the time coupling degree The calculation method includes:

[0104] Obtain the intersection point P of the predicted trajectories of the two vehicles;

[0105] Calculate the arrival time difference .

[0106] The time coupling degree can be calculated based on the following formula:

[0107] =exp * ;

[0108] Wherein τ=5 seconds is the time attenuation constant, = 3 seconds is the critical time window.

[0109] In some embodiments, the alert can include a three-level early warning mechanism:

[0110] Level 1 warning (score R of collision risk > first threshold value): dashboard visual warning + HUD risk source highlight;

[0111] Level 2 warning (R > second threshold value): increase seat vibration warning + voice prompt;

[0112] Level 3 warning (R > third threshold value): trigger active braking intervention + automatically upload event data;

[0113] And the early warning threshold is dynamically adjusted: the threshold on dry roads during the day is greater than that on rainy nights or in mountainous areas.

[0114] Through the innovation of binning rules, the risk sensitivity is improved: for example, the weight increases in the (n, n+m] interval for frequent emergency braking; data-driven optimization: dynamically update the bin boundaries based on historical data, and improve the efficiency of model iteration.

[0115] In some embodiments, based on the score card model, the WOE value of the driver's historical behavior characteristics (such as the number of illegal acts, accident records, etc.) and the model coefficient, the driver's comprehensive risk score can be calculated as the basis for interactive risk assessment:

[0116] ;

[0117] : The final benchmark score is calculated by offset (benchmark offset) and conversion coefficient, logistic regression intercept;

[0118] : WOE value of the i-th feature bin (such as "number of illegal acts of not using lights as required" bin WOE value, Table 1);

[0119] : The logistic regression model coefficient of the i-th feature;

[0120] : Recent behavior decay factor, , is the time interval (days) from the nearest behavior, (benchmark decay days, recent behavior weight is higher);

[0121] n: number of features (such as number of illegal acts, age, driving age, etc.).

[0122] In some embodiments, the interaction risk between the current vehicle and the adjacent vehicle due to the driver's behavior and the spatio-temporal coupling can also be measured, which is one of the core indicators of collision risk prediction:

[0123] ;

[0124] : scene amplification coefficient;

[0125] : current vehicle driver comprehensive score, : nearby vehicle driver comprehensive score;

[0126] : current vehicle recent score change rate, reflecting the risk rising / dropping trend;

[0127] , : time / space coupling degree dynamic weight ( , high-speed scene , ; low-speed scene vice versa);

[0128] : time coupling degree, , join the sine term, enhance the risk fluctuation sensitivity within the critical time window, seconds, seconds;

[0129] : space coupling degree, improved as , enhance the risk attenuation rate in close distance, d is the distance, meters reference distance.

[0130] In some embodiments, the impact of external environment (weather, road, visibility, etc.) on driving safety can also be evaluated, together with the interaction risk index to determine the collision risk:

[0131] ;

[0132] : weather basic risk (e.g. sunny = 0.2, rainy = 0.5, snowy = 0.8);

[0133] : visibility parameter (e.g. < 200 meters = 0.3, < 100 meters = 0.6);

[0134] : road surface state risk (e.g. dry = 0, water accumulation = 0.4, icy = 0.7, based on real-time data from roadside sensors);

[0135] : road alignment risk (e.g. straight line = 1, curve curvature > 5° / km = 1.5, sharp bend > 10° / km = 2.0);

[0136] : Field obstruction coefficient (e.g., no obstruction = 1, obstruction > 30% = 0.6, obstruction > 60% = 0.3);

[0137] : Traffic flow factor (e.g., flow < 500 vehicles / hour = 1, 500-1000 vehicles / hour = 1.3, > 1000 vehicles / hour = 1.6, based on real-time traffic flow data).

[0138] The integrated interaction risk index (IRE) is combined with the scene risk index (DSR) to determine whether there is a collision risk by quantifying the score and comparing it with the threshold value: ;

[0139] R: Integrated collision risk score;

[0140] : Time-sensitive factor (e.g., peak hours = 1.2, flat peak = 1.0, night = 1.3, dynamically adjust threshold value to adapt to period risk);

[0141] Threshold judgment: when there is a collision risk, dynamic adjustment (e.g., dry in the daytime = 8, night / rain = 6, mountainous bends = 5).

[0142] In some embodiments, the specific behavior (such as sudden braking, lane deviation, etc.) that the other party / our party may cause the collision can also be predicted based on the integrated risk score (R) and the risk factor that contributes the most:

[0143] Other party behavior prediction: combined with the comprehensive score of the driver of the nearby vehicle , recent high-risk features (such as sudden braking frequency, lane deviation proportion) and real-time data (radar detects distance change, camera identifies pressure line track), the risk factor with the highest contribution (such as and sudden braking frequency > 0.3 times / hour) is selected to predict "sudden braking".

[0144] Our party behavior prediction: combined with the comprehensive score of the current vehicle driver , real-time operation data (such as following distance < 2 seconds, visual deviation > 3 seconds) and interaction responsibility (such as rear vehicle speed > front vehicle 20 km / h, then predict "distance getting smaller").

[0145] Figure 2 A traffic assistance device structure schematic diagram based on driver traffic accident risk prediction is provided for the embodiments of the present application. As shown in Figure 2 , the device is applied to a traffic assistance system, the traffic assistance system includes a vehicle terminal, a roadside device and a server, and the device includes:

[0146] The receiving module 201 is configured to receive driving information of each vehicle, the driving information including driver information and driving data; wherein the vehicle-mounted terminal obtains the driving information and sends the driving information to the roadside device, and the roadside device reports the driving information to the server;

[0147] The obtaining module 202 is configured to obtain, from the server, a driver comprehensive score of each driver based on the driver information of each vehicle, the driver comprehensive score being determined based on a reference score and a personnel score, the reference score being determined based on an accident involvement rate, an accident frequency rate and an accident intensity rate, the accident involvement rate being a proportion of drivers involved in at least one accident in a period to total drivers, the accident frequency rate being an accident rate per person, and the accident intensity rate being a repeated number of times of an accident, and the personnel score being used to evaluate a personal driving risk score of the driver; and the driving data including space-time information and environmental data.

[0148] The predicting module 203 is configured to, for each group of adjacent vehicles, predict, based on the driving data and the driver comprehensive score of the adjacent vehicles, whether each group of the adjacent vehicles has a collision risk; and generate, for each vehicle in the adjacent vehicles having the collision risk, a collision risk event, the collision risk event including a position of an opposite vehicle having the collision risk, a predicted behavior of the opposite vehicle causing the collision and a predicted behavior of the self vehicle causing the collision.

[0149] The sending module 204 is configured to send the collision risk event to the adjacent vehicles having the collision risk.

[0150] The device embodiment corresponds to the foregoing method embodiment, and can be understood with reference to each other.

[0151] Referring to Figure 3 As shown in the figure, the electronic device 300 provided in the embodiments of the present application at least includes a processor 301, a memory 302 and a computer program stored in the memory 302 and executable on the processor 301, and the processor 301 implements the traffic assistance method based on driver traffic accident risk prediction provided in the embodiments of the present application when executing the computer program.

[0152] The electronic device 300 provided in the embodiments of the present application can further include a bus 303 connecting different components (including the processor 301 and the memory 302). Among them, the bus 303 represents one or more of several types of bus structures, including a memory bus, a peripheral bus, a local bus, etc.

[0153] The memory 302 can include a readable storage medium in the form of volatile memory, such as a random access memory (RAM) 3021 and / or cache memory 3022, and further can include a read only memory (ROM) 3023. The memory 302 can also include a program tool 3025 having a set of (at least one) program modules 3024 including, but not limited to, an operating system, one or more applications, other program modules, and program data, each of which can include implementation of a network environment, alone or in some combination.

[0154] The processor 301 can be one processing element or a collective term for a plurality of processing elements, for example, the processor 301 can be a central processing unit (CPU), or one or more integrated circuits configured to implement the driving person traffic accident risk prediction based traffic assistance method provided by the embodiments of the present application. Specifically, the processor 301 can be a general-purpose processor, including but not limited to a CPU, an application specific integrated circuit (ASIC), a ready programmable gate array (FPGA) or other programmable logic device, a discrete gate or transistor logic device, a discrete hardware component, etc.

[0155] The electronic device 300 can communicate with one or more external devices 304 (such as a keyboard, a remote control, etc.), and can also communicate with one or more devices that enable a user to interact with the electronic device 300 (such as a phone, a computer, etc.), and / or with any device that enables the electronic device 300 to communicate with one or more other electronic devices 300 (such as a router, a modem, etc.). Such communication can be carried out through an input / output (I / O) interface 305. Also, the electronic device 300 can communicate with one or more networks (such as a local area network (LAN), a wide area network (WAN), and / or a public network, such as the Internet) through a network adapter 306. As shown, the network adapter 306 communicates with other modules of the electronic device 300 through the bus 303. It should be understood that although Figure 3 the network adapter 306 is shown as a separate component from the bus 303, the network adapter 306 can be an integral part of the bus 303 or the processor 301. As shown, the bus 303 also connects the electronic device 300 to a display 307, which can be internal or external to the electronic device 300. Figure 3Other hardware and / or software modules can be used in conjunction with electronic device 300, as desired, including, but not limited to, microcode, device drivers, redundant processors, external disk drive arrays, Redundant Arrays of Independent Disks (RAID) subsystems, tape drives, data backup storage subsystems, etc., which are not shown.

[0156] It should be noted that, Figure 3 Electronic device 300 shown is merely an example and should not limit the function and use range of the embodiments of the present application.

[0157] A computer readable storage medium provided by the embodiments of the present application is introduced as follows. The computer readable storage medium provided by the embodiments of the present application stores computer instructions, and the computer instructions are executed by a processor to implement the traffic assistance method based on driver traffic accident risk prediction provided by the embodiments of the present application. Specifically, the computer instructions can be built-in or installed in the processor, so that the processor can implement the traffic assistance method based on driver traffic accident risk prediction provided by the embodiments of the present application by executing the built-in or installed computer instructions.

[0158] In addition, the traffic assistance method based on driver traffic accident risk prediction provided by the embodiments of the present application can also be implemented as a computer program product, which includes program codes that implement the traffic assistance method based on driver traffic accident risk prediction provided by the embodiments of the present application when running on a processor.

[0159] The computer program product provided by the embodiments of the present application can adopt one or more computer readable storage media, and the computer readable storage media can be, but are not limited to, an electrical, magnetic, optical, electromagnetic, infrared or semiconductor system, device or component, or any appropriate combination of the above. Specifically, more specific examples (non-exhaustive list) of the computer readable storage media include an electrical connection with one or more wires, a portable disk, a hard disk, a RAM, a ROM, an Erasable Programmable Read Only Memory (EPROM), an optical fiber, a portable Compact Disc Read-Only Memory (CD-ROM), an optical storage device, a magnetic storage device, or any appropriate combination of the above.

[0160] The computer program product provided by the embodiments of the present application can adopt a CD-ROM and include program codes, and can also run on an electronic device such as a road management device. However, the computer program product provided by the embodiments of the present application is not limited to this. In the embodiments of the present application, the computer readable storage medium can be any tangible medium containing or storing program codes, which can be used by or in combination with an instruction execution system, device or apparatus.

[0161] It should be noted that although several units or sub-units of the apparatus are mentioned in the foregoing detailed description, such division is merely exemplary and not mandatory. In fact, according to the embodiments of the present application, the features and functions of two or more units described above can be embodied in one unit. Conversely, the features and functions of one unit described above can be further divided into units for embodiment.

[0162] In addition, although the operations of the method of the present application are described in a particular order in the accompanying drawings, this does not require or imply that the operations must be performed in this particular order, or that all of the shown operations must be performed to achieve the desired result. Additionally or alternatively, certain steps can be omitted, a plurality of steps can be combined into one step, and / or one step can be divided into a plurality of steps.

[0163] Although the preferred embodiments of the present application have been described, those skilled in the art who are informed of the basic inventive concept can make additional changes and modifications to the embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications falling within the scope of the present application.

[0164] Obviously, those skilled in the art can make various modifications and variations to the embodiments of the present application without departing from the spirit and scope of the embodiments of the present application. Thus, if these modifications and variations of the embodiments of the present application fall within the scope of the claims of the present application and their equivalent technologies, the present application also intends to include these modifications and variations.

Claims

1. A traffic assistance method based on a driver traffic accident risk prediction, characterized by, The method is applied to a traffic assistance system including a vehicle terminal, a roadside device and a server, and the method comprises: The roadside device receives driving information of each vehicle, the driving information including driver information and driving data; wherein the vehicle terminal acquires the driving information and sends the driving information to the roadside device, and the roadside device reports to the server; The roadside device acquires driver comprehensive scores of each driver from the server based on the driver information of each vehicle, the driver comprehensive score being determined based on a benchmark score and a personnel score, the benchmark score being determined based on an accident involvement rate, an accident frequency rate and an accident intensity rate, the accident involvement rate being the proportion of drivers involved in at least one accident in a period to the total number of drivers, the accident frequency rate being the accident rate per person, and the accident intensity rate being the number of repeated accidents of the person, and the personnel score being used to evaluate the personal driving risk score of the driver; the driving data including space-time information and environmental data; For each group of adjacent vehicles, the roadside device predicts whether each group of adjacent vehicles has a collision risk based on the driving data and the driver comprehensive score of the adjacent vehicles; and a collision risk event is generated for each vehicle in the adjacent vehicles having a collision risk, the collision risk event including the position of the opposite vehicle having a collision risk, the predicted behavior of the opposite vehicle causing a collision and the predicted behavior of the self vehicle causing a collision; The roadside device sends the collision risk event to the adjacent vehicles having a collision risk; For each group of adjacent vehicles, predicting whether each group of adjacent vehicles has a collision risk based on the driving data and the driver comprehensive score of the adjacent vehicles comprises: For each group of adjacent vehicles, predicting the interaction risk index and the scene risk index of each group of adjacent vehicles based on the driving data and the driver comprehensive score of the adjacent vehicles; and predicting whether each group of adjacent vehicles has a collision risk based on the interaction risk index and the scene risk index of each group of adjacent vehicles. where the interaction risk index is determined based on the following formula: ； wherein, is a comprehensive score of the driver of the current vehicle; is a comprehensive score of the driver of the adjacent vehicle; is a time coupling degree , is a time difference to the conflict point, is a reference time; is a space coupling degree, , is a vehicle distance, is a reference distance; a scenario amplification coefficient; The scene risk index is determined based on the following formula: ； wherein, is a weather base risk, obtained based on meteorological applications; is a visibility parameter, obtained based on data analysis of vehicle radar point cloud; is a road alignment risk; determined based on map information; is a view obstruction coefficient, determined based on images obtained by vehicle-mounted camera.

2. The method of claim 1, wherein, The server periodically collects driving data of each driver from the roadside device, determines a benchmark score and a person score based on the driving data of each driver, the benchmark score is determined based on the following formula: ； wherein, is the accident involvement rate, is the accident frequency rate, is the accident intensity rate, is the volatility penalty coefficient, is the intensity coefficient.

3. The method of claim 2, wherein, The driver comprehensive score Score is determined based on the following formula: Score = factor coefficient WOE value; Wherein, factor is determined based on the conversion of the benchmark score, and the total WOE value is equal to the sum of the values of all key bins corresponding to the personal driving; for each key bin, WOE=ln (the proportion of high-risk drivers in the bin / the proportion of general-risk drivers in the bin); each key bin corresponds to a key feature.

4. The method of claim 3, wherein, The key features include one or more of the following: gender, whether holding an A certificate, age group, driving age group, cumulative score, number of accidents in the past three years, and number of violations in the past three years.

5. The method of claim 4, wherein, The key features further include one or more of the following: emergency braking frequency, lane deviation duration proportion, frequency of vehicle time distance less than 2s, night accident rate correction coefficient, visual line deviation time, physiological fatigue degree, and violation type.

6. The method of claim 1, wherein, The behavior of the opposite vehicle causing a collision or the behavior of the self vehicle causing a collision includes one or more of the following: Emergency braking, lane deviation, vehicle distance becoming smaller, visual line deviation, fatigue driving, and violation behavior.

7. The method of claim 6, wherein, The specific collision behavior is predicted based on the risk factor that contributes most to the interaction risk and the scene risk.

8. The method of claim 1, wherein, The vehicle terminal receives the collision risk event sent by the roadside device and gives an alarm through a vehicle voice system.

9. A traffic assistance device based on a driver traffic accident risk prediction, characterized by, The device is applied to a traffic assistance system, and the traffic assistance system comprises a vehicle terminal, a roadside device and a server. The receiving module is configured to receive driving information of each vehicle, the driving information comprising driver information and driving data; wherein the vehicle terminal acquires the driving information and sends the driving information to the roadside device, and the roadside device reports the driving information to the server; The obtaining module is configured to obtain a driver comprehensive score of each driver from the server based on the driver information of each vehicle, the driver comprehensive score being determined based on a benchmark score and a personnel score, the benchmark score being determined based on an accident involvement rate, an accident frequency rate and an accident intensity rate, the accident involvement rate being a proportion of drivers involved in at least one accident in a period to total drivers, the accident frequency rate being an accident rate per person, and the accident intensity rate being a repeated number of accidents of a person, and the personnel score being used to evaluate a personal driving risk score of the driver; and the driving data comprising space-time information and environmental data. The prediction module is configured to, for each group of adjacent vehicles, predict whether each group of the adjacent vehicles has a collision risk based on the driving data and the driver comprehensive score of the adjacent vehicles; and generate a collision risk event for each vehicle in the adjacent vehicles having the collision risk, the collision risk event comprising a position of an opposite vehicle having the collision risk, a predicted behavior of the opposite vehicle causing a collision and a predicted behavior of the self causing a collision. The sending module is configured to send the collision risk event to the adjacent vehicles having the collision risk. The prediction module is further configured to, for each group of adjacent vehicles, predict an interaction risk index and a scene risk index of each group of the adjacent vehicles based on the driving data and the driver comprehensive score of the adjacent vehicles; and predict whether each group of the adjacent vehicles has a collision risk based on the interaction risk index and the scene risk index of each group of the adjacent vehicles. where the interaction risk index is determined based on the following formula: ； wherein, is a comprehensive score of the driver of the current vehicle; is a comprehensive score of the driver of the proximate vehicle; is a time coupling degree , is a time difference to reach the conflict point, is a reference time; is a space coupling degree, , is a vehicle distance, is a reference distance; a scenario amplification coefficient; The scene risk index is determined based on the following formula: ； wherein, is a weather base risk, obtained based on meteorological applications; is a visibility parameter, obtained based on data analysis of vehicle radar point cloud; is a road alignment risk; determined based on map information; is a view obstruction coefficient, determined based on images obtained by vehicle-mounted cameras.

Images