A quantitative assessment method for motor vehicle accident risk at intersections

Abstract

The present invention provides a method for quantitatively assessing the risk of motor vehicle accidents at intersections, including: obtaining historical vehicle trajectory data within the intersection influence area and preprocessing to obtain preprocessed data; extracting vehicle dangerous driving behavior feature data including sudden acceleration and deceleration from the preprocessed data; respectively calculating the average daily frequency of dangerous vehicle behaviors under different turning labels; statistically calculating the average daily motor vehicle trajectory flow under different turning labels based on the vehicle historical trajectory data; calculating the spatiotemporal distribution of motor vehicle accident risk probabilities under different turning labels and generating a spatiotemporal map; and weighted calculating the overall motor vehicle accident risk level index at the intersection. The present invention utilizes the frequency characteristics of dangerous driving behaviors of networked vehicles to correlate with intersection accident probability characteristics, effectively making up for the difficulty in collecting traffic accident data, and effectively supporting safety warning services for vehicles passing through intersections.

Description

Technical Field

[0001] The invention relates to the field of traffic management, and in particular to a method for quantitatively assessing the risk of motor vehicle accidents at intersections. Background Art

[0002] With the continuous growth of motor vehicle travel, motor vehicle accidents at urban intersections are also increasing rapidly. Quantitative assessment and early warning services of traffic safety at intersections have become an important part of traffic management.

[0003] Existing technical solutions for quantitative evaluation of intersection accident risks usually use characteristic parameters such as vehicle collision time (TTC) and relative speed (i.e., the speed difference between the front and rear vehicles) during vehicle following as key indicators to model and predict vehicle and even intersection accident risks. This solution requires real-time (sampling frequency is at least 1 time / second) collection of high-frequency data such as the motion state and relative position information of vehicles in the intersection area. There are also solutions that use historical vehicle accident data to analyze the correlation between intersection-related factors (such as road geometry, slope, speed limit, etc.) and traffic accidents, so as to statistically analyze and predict the probability of vehicle accidents at intersections.

[0004] Existing technical solutions often have high requirements for data collection and are difficult to apply in engineering. Both micro-risk assessment based on car-following behavior parameters and macro-risk assessment based on accidents / traffic violations have high requirements for data collection and coverage quality, which makes it difficult to apply the technology on a large scale. Although the gradually developed video technology has made up for this disadvantage to a certain extent, it is still difficult to cover a large number of intersections and large-scale spaces upstream and downstream of intersections.

[0005] The risk assessment of existing solutions usually focuses on the entire intersection, with a coarse granularity and insufficient precision. At the same time, existing technologies often lack consideration of the differences in the temporal and spatial risks of intersections, and cannot reflect the differences in risk characteristics at different spatial locations and under different time conditions of intersections, resulting in a relatively rough and low-precision safety warning service for supporting vehicles. Summary of the invention

[0006] In view of the above problems, the present invention is proposed to provide a method for quantitatively assessing the risk of motor vehicle accidents at intersections that overcomes the above problems or at least partially solves the above problems.

[0007] According to one aspect of the present invention, a method for quantitatively assessing the risk of motor vehicle accidents at intersections is provided, comprising:

[0008] Obtaining historical trajectory data of vehicles in the intersection influence area and preprocessing them to obtain preprocessed data;

[0009] Extracting vehicle dangerous driving behavior feature data including sudden acceleration and sudden deceleration from the preprocessed data;

[0010] Calculate the average daily frequency of dangerous vehicle behaviors under different turning labels;

[0011] Based on the vehicle historical trajectory data, the daily average vehicle trajectory flow under different turning labels is counted;

[0012] Calculate the spatiotemporal distribution of motor vehicle accident risk probability under different turn labels and generate a spatiotemporal map;

[0013] The overall motor vehicle accident risk level index of the intersection is calculated by weighted calculation.

[0014] Optionally, the obtaining and preprocessing of historical vehicle trajectory data in the intersection influence area specifically includes:

[0015] For different directions of the intersection, the intersection impact area is defined as a certain distance from the upstream and downstream sections, with the stop line of the intersection entrance as the center;

[0016] Extracting historical vehicle trajectory data within the intersection influence area according to the latitude and longitude information in the vehicle historical trajectory data;

[0017] The trajectory is divided into different data sets according to the two labels of the entrance road direction and the trajectory turning, and different vehicle turning labels are assigned;

[0018] Check the continuity of the data time based on the second-by-second trajectory data and calculate the instantaneous acceleration of the vehicle. The calculation formula is:

[0019]

[0020] Among them: a i For vehicles at t i The instantaneous acceleration at the moment m / s 2 , v i For vehicles at t i Instantaneous speed at time t m / s, i is the timestamp corresponding to the instantaneous speed, and i is a positive integer.

[0021] Optionally, the distance of the upstream section is 500 meters, and the distance of the downstream section is 300 meters.

[0022] Optionally, extracting the vehicle dangerous driving behavior characteristic data including sudden acceleration and sudden deceleration from the preprocessed data specifically includes:

[0023] Extract the sudden acceleration and deceleration behaviors from the intersection vehicle trajectory data respectively; the diagnosis rule of sudden acceleration: the speed and acceleration changes in the process must satisfy the following formula:

[0024]

[0025]

[0026] Among them: a a is the critical acceleration of the vehicle during rapid acceleration m / s 2 , m and n are positive integers;

[0027] Rapid deceleration diagnostic rules: Speed ​​and acceleration changes must satisfy the following formula:

[0028]

[0029]

[0030] Among them: a d is the critical acceleration of the vehicle's rapid deceleration m / s 2 , m and n are positive integers;

[0031] The dangerous driving behaviors extracted from the trajectory are numbered in sequence, and the vehicle turning label, the occurrence time and location data of the dangerous driving behavior are extracted respectively.

[0032] Optionally, the calculating the average daily occurrence frequency of dangerous vehicle behaviors under different turning labels specifically includes:

[0033] The time-space statistical interval of the intersection is divided into fixed time intervals I and distance intervals D;

[0034] Based on the extracted dangerous driving behavior data, the frequency of dangerous vehicle behaviors under different steering, time periods and spatial location conditions is calculated.

[0035] represents the sum of the daily average number of sudden accelerations and decelerations of motor vehicles at time period t and spatial position l under the intersection vehicle turning label k;

[0036] The set of spatiotemporal statistical intervals of the intersection is represented as (k, t, l) ∈ {K, T, L}, where K is the vehicle turn label set, T and L are the time period set and spatial location set of the intersection, respectively.

[0037] Optionally, the counting of daily average motor vehicle trajectory flow under different turning labels based on the vehicle historical trajectory data specifically includes:

[0038] Based on the historical vehicle trajectories, the daily average vehicle trajectory flow under different turning conditions during the statistical period is calculated. It represents the average daily trajectory flow of motor vehicles in time period t under the condition of turning k at the intersection.

[0039] Optionally, the calculating the spatiotemporal distribution of the motor vehicle accident risk probability under different turn labels and generating a spatiotemporal map specifically includes:

[0040] First, calculate the equivalent violation frequency of each spatiotemporal position (t, l) under different turn labels. The calculation formula is:

[0041]

[0042] in: It represents the frequency of equivalent dangerous driving behaviors of motor vehicles at time t and spatial position l under the intersection turn label k during the statistical period / veh / D;

[0043] Calculate the spatiotemporal distribution of the vehicle accident risk probability under the intersection turn k label according to the equivalent violation frequency The calculation formula is:

[0044]

[0045] According to the spatiotemporal distribution Draw a spatiotemporal map of intersection accident risk under the intersection turn k label.

[0046] Optionally, the weighted calculation of the overall motor vehicle accident risk level index of the intersection specifically includes:

[0047] First, filter out the maximum risk probability under each turn label k, and the calculation formula is:

[0048]

[0049] Determine the weight value g under each turning label k based on the average daily motor vehicle trajectory flow k , the calculation formula is:

[0050]

[0051] According to the weight value g k The overall accident risk index R of the intersection is weighted and calculated using the following formula:

[0052]

[0053] The present invention provides a method for quantitatively assessing the risk of motor vehicle accidents at intersections, including: obtaining historical vehicle trajectory data within the intersection influence area and preprocessing it to obtain preprocessed data; extracting vehicle dangerous driving behavior feature data including sudden acceleration and deceleration from the preprocessed data; respectively calculating the average daily frequency of dangerous vehicle behaviors under different turning labels; statistically analyzing the average daily motor vehicle trajectory flow under different turning labels based on the vehicle historical trajectory data; calculating the spatiotemporal distribution of motor vehicle accident risk probabilities under different turning labels and generating a spatiotemporal map; and weightedly calculating the overall motor vehicle accident risk level index at the intersection. The present invention utilizes the frequency characteristics of dangerous driving behaviors of networked vehicles to correlate with intersection accident probability characteristics, thereby achieving high-resolution quantitative assessment of the level of vehicle accident risk lines at intersections, effectively making up for the difficulty in collecting traffic accident data; while improving the spatiotemporal coverage and quantitative accuracy of intersection accident risk assessment, it also effectively supports safety warning services for vehicles passing through intersections.

[0054] The above description is only an overview of the technical solution of the present invention. In order to more clearly understand the technical means of the present invention, it can be implemented according to the contents of the specification. In order to make the above and other purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0056] Figure 1 A flowchart of a method for quantitatively assessing the risk of motor vehicle accidents at intersections provided by an embodiment of the present invention;

[0057] Figure 2 An example schematic diagram of the intersection influence area from south to north provided in an embodiment of the present invention;

[0058] Figure 3 A spatiotemporal diagram of intersection accident risks under the south entrance straight-ahead condition provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0059] The exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

[0060] The terms "comprises" and "having" and any variations thereof in the description embodiments, claims and drawings of the present invention are intended to cover non-exclusive inclusions, for example, including a series of steps or units.

[0061] The technical solution of the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments.

[0062] like Figure 1 As shown, step 1) extracts historical vehicle trajectory data within the upstream and downstream spatial range of the intersection as the center and performs preprocessing.

[0063] For different directions of the intersection, the center is the stop line of the intersection entrance lane, such as Figure 2 As shown in the figure, the length of the upstream section is 500 meters and the length of the downstream section is 300 meters as the intersection influence area. The upstream starting point, stop line and downstream end point are marked as -500m, 0m and 300m respectively.

[0064] The distance between the upstream and downstream sections can be flexibly determined according to the characteristics of the upstream and downstream of the intersection; but the upstream is generally not shorter than 300 meters and the downstream is not shorter than 150 meters.

[0065] The data in the intersection influence area is extracted based on the latitude and longitude information in the historical vehicle trajectory data, and divided into different data sets based on the two labels of the entrance road direction and trajectory turning, and different vehicle turning labels k are assigned. Figure 2 As shown, at a standard cross intersection, there are 12 vehicle turning labels in total, as shown in Table 1 below:

[0066] Table 1 Vehicle turn label set at cross intersections

[0067] Go straight from the north entrance Turn left at the north entrance Turn right at the north entrance Go straight from the south entrance Turn left at the South Entrance Turn right at the south entrance East entrance straight Turn left at the east entrance Turn right at the east entrance Go straight from the west entrance Turn left at the west entrance Turn right at the west entrance

[0068] Then the data time continuity is checked for the second-by-second trajectory data, and the data with time errors, speed, and missing longitude and latitude information are eliminated. Then the instantaneous acceleration of the vehicle is calculated using the following formula:

[0069]

[0070] Among them: a i For vehicles at t i The instantaneous acceleration at the moment (m / s 2 )v i For vehicles at t i Instantaneous speed at time t (m / s), i is the timestamp corresponding to the instantaneous speed, and i is a positive integer.

[0071] Step 2) According to the dangerous driving behavior diagnosis rules, diagnose the rapid acceleration and rapid deceleration behaviors in the trajectory respectively. The diagnosis rule of rapid acceleration must satisfy the formula:

[0072]

[0073]

[0074] Among them: a a is the critical acceleration of the vehicle during rapid acceleration (m / s 2 ), m and n are positive integers.

[0075] The rapid deceleration diagnostic rule must satisfy the formula:

[0076]

[0077]

[0078] Among them: a d is the critical acceleration of the vehicle's rapid deceleration (m / s 2 ), m and n are positive integers.

[0079] The dangerous driving behaviors (including sudden acceleration and deceleration) extracted from the trajectory are numbered in sequence, and the vehicle turning label, the occurrence time of the dangerous driving behavior (based on the start time of sudden acceleration or deceleration, accurate to minutes) and the location data (latitude and longitude information) are extracted respectively. The detailed information is shown in Table 2.

[0080] Table 2 Example of dangerous driving behavior data of vehicles at cross intersections

[0081]

[0082] Step 3) Based on the extracted dangerous driving behavior data, the statistical results of dangerous driving behavior and traffic volume during the statistical period are calculated. The sample data is shown in Table 3. The statistical interval is divided into fixed time intervals I = 30 minutes and distance intervals D = 50 meters. The frequency of dangerous behaviors of vehicles under different turning, different spatial and temporal conditions is calculated respectively. It represents the sum of the daily average number of sudden acceleration and deceleration times of motor vehicles in the T period and L spatial position under the vehicle turning label k at the intersection (in times / I / D).

[0083] The fixed time interval I can be a time interval of 5 minutes and its integer multiples (such as 10 minutes, 15 minutes, 60 minutes, etc.), and the fixed distance interval D can be a distance interval of 10 meters and its integer multiples (such as 30 meters, 50 meters, 100 meters); the specific time interval and distance interval values ​​can be determined according to the situation of the intersection.

[0084] Step 4) Calculate the daily average vehicle trajectory flow of different turning conditions during the statistical period based on the vehicle historical trajectory As shown in Table 3. It represents the average daily trajectory flow of motor vehicles in time period t under the condition of turning K at the intersection (in veh / I).

[0085] The daily average trajectory flow proposed by the present invention only counts the number of vehicles with trajectory records in the trajectory database, so the trajectory flow is not equivalent to the intersection cross-sectional flow. At the same time, the statistical period of the sample data should not be less than 20 natural days.

[0086] Table 3. Example of statistics on the frequency of dangerous driving behaviors and daily average trajectory flow of vehicles

[0087]

[0088] Step 5) calculates the spatiotemporal distribution of the motor vehicle accident risk probability under the intersection turn k label according to the statistical results of step 3) and step 4).

[0089] First, the equivalent violation frequency of each spatiotemporal position (t, l) under different turn labels is calculated. The sample results are shown in Table 3. The calculation formula is:

[0090]

[0091] in: It represents the frequency of equivalent dangerous driving behaviors of motor vehicles under intersection turn label k at time period t and spatial position l during the statistical period (times / veh / D).

[0092] According to the results of the equivalent violation frequency, the spatiotemporal distribution of the vehicle accident risk probability under the intersection turn k label is further calculated The calculation formula is:

[0093]

[0094] Based on the above statistical results, a spatiotemporal diagram of the intersection accident risk under the intersection turn k label is drawn. For example, Figure 3 shown.

[0095] Step 6) weightedly calculates the overall motor vehicle accident risk level index of the intersection according to the calculation results of step 5).

[0096] First, the maximum risk probability under each turn label k is screened out, as shown in Table 4. The calculation formula is:

[0097]

[0098] Determine the weight value g under each turning label k based on the average daily motor vehicle trajectory flow k, an example is shown in Table 4. The calculation formula is:

[0099]

[0100] The overall accident risk index R of the intersection is calculated based on the weight value, and the results are shown in Table 4. The calculation formula is:

[0101]

[0102] Table 4 Example of calculation of vehicle accident risk index at intersection

[0103]

[0104] Beneficial effects: The present invention utilizes the frequency characteristics of dangerous driving behaviors of networked vehicles to correlate with the probability characteristics of intersection accidents, achieving high-resolution quantitative assessment of the level of vehicle accident risk lines at intersections, effectively compensating for the difficulty in collecting traffic accident data; while improving the spatiotemporal coverage and quantitative accuracy of intersection accident risk assessment, it also effectively supports safety warning services for vehicles passing through intersections.

[0105] It overcomes the difficulty of collecting relevant data, and at the same time has the ability to quantify the temporal and spatial differences in intersection accident risks, supporting precise safety warning services.

[0106] The above specific implementation methods further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above are only specific implementation methods of the present invention and are not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for quantitatively assessing the risk of motor vehicle accidents at intersections, characterized in that: The evaluation methods include: Obtain the historical trajectory data of vehicles in the intersection influence area and preprocess it to obtain preprocessed data, including: For different directions of the intersection, the intersection impact area is defined as a certain distance from the upstream and downstream sections, with the stop line of the intersection entrance as the center; Extracting historical vehicle trajectory data within the intersection influence area according to the latitude and longitude information in the vehicle historical trajectory data; The trajectory is divided into different data sets according to the two labels of the entrance road direction and the trajectory turning, and different vehicle turning labels are assigned; Check the continuity of the data time based on the second-by-second trajectory data and calculate the instantaneous acceleration of the vehicle. The calculation formula is: Among them: a i For vehicles at t i The instantaneous acceleration at the moment m / s 2 , v i For vehicles at t i Instantaneous speed at time t m / s, i is the timestamp corresponding to the instantaneous speed, i is a positive integer; The upstream section distance is 500 meters, and the downstream section distance is 300 meters; Extracting vehicle dangerous driving behavior feature data including sudden acceleration and sudden deceleration from the preprocessed data; Calculate the average daily frequency of dangerous vehicle behaviors under different turning labels, including: The time-space statistical interval of the intersection is divided into fixed time intervals I and distance intervals D; Based on the extracted dangerous driving behavior data, the frequency of dangerous vehicle behaviors under different steering, time periods and spatial location conditions is calculated. represents the sum of the daily average number of sudden accelerations and decelerations of motor vehicles at time period t and spatial position l under the intersection vehicle turning label k; The set of spatiotemporal statistical intervals of the intersection is represented as (k, t, l) ∈ {K, T, L}, where K is the vehicle turn label set, T and L are the time period set and spatial location set of the intersection respectively; Based on the vehicle historical trajectory data, the daily average vehicle trajectory flow under different turning labels is counted; Calculate the spatiotemporal distribution of motor vehicle accident risk probability under different turn labels and generate a spatiotemporal map; The overall motor vehicle accident risk level index of the intersection is calculated by weighted calculation.

2. The method for quantitatively assessing the risk of motor vehicle accidents at intersections according to claim 1, characterized in that: The extracting of the vehicle dangerous driving behavior characteristic data including sudden acceleration and sudden deceleration from the preprocessed data specifically includes: Extract the sudden acceleration and deceleration behaviors from the intersection vehicle trajectory data respectively; the diagnosis rule of sudden acceleration: the speed and acceleration changes in the process must satisfy the following formula: Among them: a a is the critical acceleration of the vehicle during rapid acceleration m / s 2 , m and n are positive integers; Rapid deceleration diagnostic rules: Speed ​​and acceleration changes must satisfy the following formula: Among them: a d is the critical acceleration of the vehicle's rapid deceleration m / s 2 , m and n are positive integers; The dangerous driving behaviors extracted from the trajectory are numbered in sequence, and the vehicle turning label, the occurrence time and location data of the dangerous driving behavior are extracted respectively.

3. The method for quantitatively assessing the risk of motor vehicle accidents at intersections according to claim 1, characterized in that: The statistical method of the daily average motor vehicle trajectory flow under different turning labels based on the vehicle historical trajectory data specifically includes: Based on the historical vehicle trajectories, the daily average vehicle trajectory flow under different turning conditions during the statistical period is calculated. It represents the average daily trajectory flow of motor vehicles in time period t under the condition of turning k at the intersection.

4. The method for quantitatively assessing the risk of motor vehicle accidents at intersections according to claim 1, characterized in that: The step of calculating the spatiotemporal distribution of the motor vehicle accident risk probability under different turn labels and generating a spatiotemporal map specifically includes: First, calculate the equivalent violation frequency of each spatiotemporal position (t, l) under different turn labels. The calculation formula is: in: It represents the frequency of equivalent dangerous driving behaviors of motor vehicles at time period t and spatial position l under intersection turn label k during the statistical period, in units of times / veh / D; Calculate the spatiotemporal distribution of the vehicle accident risk probability under the intersection turn k label according to the equivalent violation frequency The calculation formula is: According to the spatiotemporal distribution Draw a spatiotemporal map of intersection accident risk under the intersection turn k label.

5. The method for quantitatively assessing the risk of motor vehicle accidents at intersections according to claim 1, characterized in that: The weighted calculation of the overall motor vehicle accident risk level index of the intersection specifically includes: First, filter out the maximum risk probability under each turn label k, and the calculation formula is: Determine the weight value g under each turning label k based on the average daily motor vehicle trajectory flow k , the calculation formula is: According to the weight value g k The overall accident risk index R of the intersection is weighted and calculated using the following formula:

Images