Expressway traffic safety state assessment method and system

Abstract

The invention relates to the technical field of highway traffic safety assessment, and provides a highway traffic safety state assessment method and system. Comprising the following steps: acquiring highway historical accident data and road alignment data, and matching the accident data with the road alignment data; the accident rate under each time index and the road alignment index is counted, and a time-space risk index scoring standard is formulated based on the accident rate; calculating the subjective weight of the time index by adopting an analytic hierarchy process, and calculating the objective weight of the road alignment index by adopting an entropy weight method; and constructing a space-time risk index model by integrating the risk score of each index and the weight coefficient, and clustering the space-time risk index based on a K-means clustering algorithm, and dividing the space-time risk index into four safety risk levels. According to the method, the highway traffic safety state can be accurately evaluated, the accident high-incidence period and the accident frequent-incidence period can be identified, it is ensured that targeted management and control measures are formulated at the first time when the high-risk state occurs in the road traffic, and therefore the highway traffic safety level is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of highway traffic safety assessment, and in particular to a method and system for assessing highway traffic safety status. Background Art

[0002] As an important part of the modern transportation network, the traffic safety status of expressways is directly related to the safety of life and property of road users. Relevant studies have shown that factors such as the terrain and road conditions of expressways have a significant impact on traffic safety, and poor road alignment is the main cause of traffic accidents. In order to improve the traffic safety level of expressways, it is particularly important to build a scientific and reasonable traffic safety assessment system.

[0003] Traditional evaluation methods are mostly based on historical traffic accident data and road alignment data, and evaluate traffic safety conditions through statistical analysis or model fitting. However, these methods often ignore the time factor of traffic accidents and are difficult to fully reflect the traffic safety status of highways under different time and space conditions.

[0004] In view of this, there is an urgent need to design a highway traffic safety status assessment method and system. Summary of the invention

[0005] The purpose of the present invention is to provide a method and system for evaluating the traffic safety status of a highway. First, historical accident data and road alignment data of the highway are obtained, the accident data are matched with the road alignment data to obtain processed data, the accident rate under each time index and road alignment index is statistically calculated, and a scoring standard for spatiotemporal risk index is formulated based on the accident rate; then, a hierarchical analysis method is used to calculate the subjective weight of the time index, and an entropy weight method is used to calculate the objective weight of the road alignment index; finally, a spatiotemporal risk index model is constructed by comprehensively combining the risk scores and weight coefficients of each index, and the spatiotemporal risk index is clustered based on the K-means clustering algorithm to divide it into four safety risk levels.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] In a first aspect, the present application provides a method for evaluating the traffic safety status of a highway, comprising the following steps:

[0008] S1: Obtain historical highway accident data and road alignment data from relevant departments, match the accident data with the road alignment data, and obtain processed data;

[0009] S2: Count the accident rates under various time indicators and road linear indicators, and formulate scoring standards for spatiotemporal risk indicators based on the accident rates;

[0010] S3: The subjective weight of the time index is calculated by using the analytic hierarchy process, and the objective weight of the road alignment index is calculated by using the entropy weight method;

[0011] S4: The spatiotemporal risk index model is constructed by integrating the risk scores and weight coefficients of each indicator and calculating the spatiotemporal risk index of the road;

[0012] S5: Cluster the spatiotemporal risk index based on the K-means clustering algorithm and divide it into four security risk levels.

[0013] Furthermore, historical accident data and road alignment data of the expressway are obtained. The accident data include: date, day of the week, time and pile number of the accident; the road alignment data include: road plane alignment data and longitudinal section alignment data.

[0014] Furthermore, the accident data is matched with the road alignment data, data with incomplete information records are eliminated, and the horizontal and vertical alignment indicators of the road at the accident location are obtained through the road alignment data according to the pile number where the accident occurred.

[0015] Furthermore, a scoring standard for spatiotemporal risk indicators is formulated based on the accident rate: an accident rate of 0-10% is low risk, with a score of 1; an accident rate of 10-15% is medium risk, with a score of 2; an accident rate of 15-20% is relatively high risk, with a score of 3; and an accident rate of more than 20% is high risk, with a score of 4.

[0016] Furthermore, the analytic hierarchy process is used to calculate the weight of each time indicator, which includes: month, week, and time; the entropy weight method is used to calculate the weight of each road alignment indicator, which includes: straight line length, circular curve radius, curve deflection angle, uphill slope, downhill slope, slope length, and vertical curve radius.

[0017] Furthermore, the temporal and spatial risk index (TSRI) model is constructed by integrating the risk scores and weight coefficients of each indicator. The calculation formula is as follows:

[0018] TSRI=w TRI TRI+w ARI ARI

[0019] Where: TSRI is the spatiotemporal risk index; TRI is the time risk index; ARI is the linear risk index; w k is the weight of the risk item in TSRI, k = {TRI, ARI};

[0020] The TRI calculation formula is as follows:

[0021] TRI=[T 1 T 2 T 3 ]×[S 1 S 2S 3 ] T

[0022] Where: T i is the risk level under the i-th time index; S i is the weight of the i-th time indicator;

[0023] The ARI calculation formula is as follows:

[0024] ARI=[X 1 X 2 X 3 X 4 X 5 X 6 X 7 ]×

[0025] [ω 1 ω 2 ω 3 ω 4 ω 5 ω 6 ω 7 ] T

[0026] Where: X i is the risk level of the th linear indicator; ω i is the weight of the i-th linear indicator;

[0027] The weight of each risk item in the spatiotemporal risk index is calculated as follows:

[0028]

[0029] Where: k is the sample standard deviation of the time risk index or linear risk index; k = {TRI, ARI}.

[0030] Furthermore, the spatiotemporal risk index is clustered based on the K-means clustering algorithm and divided into four security risk levels:

[0031] Set the number of clusters k = 4, randomly select 4 initial centroids, calculate the distance from each data point to the 4 centroids, and assign each point to the nearest centroid to form 4 clusters. Recalculate the centroid of each cluster, that is, calculate the average of all points in each cluster as the new centroid, and check whether the change of the centroid is less than a certain threshold or reaches the maximum number of iterations. If not, repeat the assignment and update steps until the centroid position of each cluster no longer moves. According to the clustering threshold, the spatiotemporal risk index is divided into 4 safety risk levels: low risk, medium risk, higher risk and high risk.

[0032] On the other hand, the present application provides a highway traffic safety status assessment system, the system comprising:

[0033] The data acquisition module is used to obtain historical accident data and road alignment data of expressways; the accident data includes: date, week, time and pile number of the accident; the road alignment data includes: road plane alignment data and longitudinal section alignment data.

[0034] The data processing module uses data cleaning and data matching methods to pre-process redundant data, missing data, abnormal data and key data from different data sources to obtain processed data; then, the temporal and spatial distribution characteristics of highway traffic accidents are analyzed through descriptive statistical methods, and a temporal and spatial risk index scoring standard is formulated based on the accident rate. An accident rate of 0-10% is low risk, with a score of 1; an accident rate of 10-15% is medium risk, with a score of 2; an accident rate of 15-20% is relatively high risk, with a score of 3; an accident rate of more than 20% is high risk, with a score of 4;

[0035] Indicator weight calculation module, the time indicator selects three indicators: month, week, and time; the road alignment indicator selects seven indicators: straight line length, circular curve radius, curve deflection angle, uphill slope, downhill slope, slope length, and vertical curve radius, to build a spatiotemporal risk indicator system; the hierarchical analysis method is used to calculate the subjective weight of the time indicator, and the entropy weight method is used to calculate the objective weight of the road alignment indicator;

[0036] The traffic status assessment module constructs a spatiotemporal risk index model based on the risk scores and weight coefficients of each indicator, and clusters the spatiotemporal risk index based on the K-means clustering algorithm, dividing it into four safety risk levels: low risk, medium risk, higher risk and high risk; calculates the real-time spatiotemporal risk index, and determines the real-time traffic safety status according to the thresholds of each risk level.

[0037] The beneficial effects of the present invention include:

[0038] Obtain historical accident data and road alignment data of highways, and match accident data with road alignment data; Count the accident rates under various time indicators and road alignment indicators, and formulate a scoring standard for spatiotemporal risk indicators based on the accident rates; Use the analytic hierarchy process to calculate the subjective weight of the time indicator, and use the entropy weight method to calculate the objective weight of the road alignment indicator; Construct a spatiotemporal risk index model based on the risk scores and weight coefficients of each indicator, and cluster the spatiotemporal risk index based on the K-means clustering algorithm, and divide it into four safety risk levels. By accurately evaluating the traffic safety status of highways, the present invention can timely discover and identify accident-prone periods and accident-prone sections. This helps relevant departments to formulate targeted control measures as soon as a high-risk state occurs, thereby effectively reducing the accident rate and improving the traffic safety level of highways. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 A flow chart of a highway traffic safety status assessment method provided by the present invention;

[0040] Figure 2 A schematic diagram of a highway traffic safety status assessment system provided by the present invention;

[0041] Figure 3 It is a variation trend diagram of the spatiotemporal risk index of each road section in a specific implementation manner of the present invention;

[0042] Figure 4 It is a distribution diagram of the spatiotemporal risk index in different months in a specific implementation manner of the present invention;

[0043] Figure 5 It is a distribution diagram of the spatiotemporal risk index in different weeks in a specific implementation manner of the present invention;

[0044] Figure 6 It is a distribution diagram of the spatiotemporal risk index in different time periods in a specific implementation manner of the present invention; DETAILED DESCRIPTION

[0045] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings of the present invention specification. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other implementation methods obtained by ordinary technicians in the field without creative work are within the scope of protection of the present application.

[0046] The embodiment of the present invention provides a method for evaluating the traffic safety status of a highway. Figure 1 As shown, the following will refer to Figure 1 The specific implementation steps of the method of the present invention are described.

[0047] S101: Obtain historical highway accident data and road alignment data from relevant departments, match the accident data with the road alignment data, and obtain processed data;

[0048] In this specific implementation, the Hebei section of the Beijing-Harbin Expressway is selected as the test section. The section is 200 km long and is located in a plain and hilly area. The design speed is 120 km / h and there are six lanes in both directions. The original traffic accident data and road alignment design drawings from 2017 to 2022 are obtained through relevant departments. The accident data include: accident number, accident time, accident location, accident severity, etc. The road alignment design drawings include plane design drawings and longitudinal section design drawings. By sorting out the original archival materials of highway traffic accidents and eliminating data with incomplete information records, a total of 6,205 traffic accident data on the main line of the section were finally collected. Through the road alignment design drawing, the horizontal and vertical alignment indicators of the road at the accident location can be obtained. The accident data is matched with the road alignment data, and the processed part of the data is shown in Table 1.

[0049] Table 1 Processed data

[0050]

[0051] S102: Counting the accident rates under various time indicators and road linear indicators, and formulating a scoring standard for the spatiotemporal risk indicator based on the accident rates;

[0052] By conducting statistical analysis on the processed traffic accident data, we can obtain the accident rates under different time dimensions and various linear indicators. Taking the accident rate as the risk level classification standard, we formulate the scoring standards for various time indicators and linear indicators from the perspective of time and space: an accident rate of 0-10% is low risk, with a score of 1; an accident rate of 10-15% is medium risk, with a score of 2; an accident rate of 15-20% is relatively high risk, with a score of 3; an accident rate of more than 20% is high risk, with a score of 4. The value range and scoring standards of each indicator are shown in Table 2.

[0053] Table 2 Time and linearity index scoring standards

[0054]

[0055] S103: using the analytic hierarchy process to calculate the subjective weight of the time index, and using the entropy weight method to calculate the objective weight of the road alignment index;

[0056] The analytic hierarchy process is widely used in the field of safety evaluation. Its basic idea is to decompose the relevant elements of a complex decision-making problem into several levels according to needs, and then use a certain scale to compare and assign values ​​to the elements in the level, objectively quantify people's subjective judgment, and thus obtain the weights in the evaluation system. The analytic hierarchy process (AHP) is selected to calculate the weights of each time indicator. The calculation results of the weights of each time indicator are shown in Table 3.

[0057] Table 3 Weights of various time indicators

[0058] index month Week time Weight 0.1172 0.2684 0.6144

[0059] The entropy weight method is an objective weighting method that uses information entropy to determine the weights of each evaluation index, avoiding the subjectivity of artificial weighting, making the weight distribution more reasonable and scientific, and able to reflect the actual role of each index in the comprehensive evaluation. The specific calculation steps for calculating the weights of road linear indicators using the entropy weight method are as follows:

[0060] Step 1: Construct the decision matrix R based on the original data.

[0061] Suppose there are n objects to be evaluated, each object has m indicators, and the jth evaluation indicator value of the i-th evaluation object is r ij , the original data matrix R is obtained as:

[0062]

[0063] Step 2: Data standardization.

[0064] Positive indicator normalization:

[0065]

[0066] Reverse indicator normalization:

[0067]

[0068] Construct the normalized matrix Y:

[0069]

[0070] Step 3: Calculate the entropy value.

[0071] First, calculate the characteristic proportion p ij , which represents the probability that the i-th value of the j-th indicator causes the entropy value of the entire evaluation system to change.

[0072]

[0073] Then calculate the entropy value:

[0074]

[0075] Step 4: Calculate the weights.

[0076] First, calculate the coefficient of variation d i , which represents the discrete degree of all evaluation values ​​of the i-th indicator, and the calculation formula is as follows:

[0077] d i =1-e i

[0078] Then calculate the evaluation index weights:

[0079]

[0080] The calculation results of the weights of each road alignment index are shown in Table 4.

[0081] Table 4 Weights of various road alignment indicators

[0082] index Weight Straight line length 0.1307 Circular curve radius 0.2509 Curve deflection 0.0940 Uphill slope 0.0879 Downhill slope 0.1624 Slope length 0.1681 Vertical curve radius 0.1060

[0083] S104: constructing a spatiotemporal risk index model by integrating the risk scores and weight coefficients of various indicators and calculating the spatiotemporal risk index of the road;

[0084] The calculation formula of the Time-Space Risk Index (TSRI) is as follows:

[0085] TSRI=w TRI TRI+w ARI ARI

[0086] Where: TSRI is the spatiotemporal risk index; TRI is the time risk index; ARI is the linear risk index; w k is the weight of the risk item in TSRI, k = {TRI, ARI};

[0087] The TRI calculation formula is as follows:

[0088] TRI=[T 1 T 2 T 3 ]×[S 1 S 2 S 3 ] T

[0089] Where: T i is the risk level under the i-th time index; S i is the weight of the i-th time indicator;

[0090] The ARI calculation formula is as follows:

[0091] ARI=[X 1 X 2 X 3 X 4 X 5 X 6 X 7 ]×

[0092] [ω 1 ω 2 ω 3 ω 4 ω 5 ω 6 ω 7 ] T

[0093] Where: X i is the risk level of the th linear indicator; ω i is the weight of the i-th linear indicator;

[0094] The weight of each risk item in the spatiotemporal risk index is calculated as follows:

[0095]

[0096] Where: k is the sample standard deviation of the time risk index or linear risk index; k = {TRI, ARI}.

[0097] S105: Clustering the spatiotemporal risk index based on the K-means clustering algorithm and dividing it into four safety risk levels.

[0098] K-means clustering algorithm is a commonly used research method in current traffic safety analysis. K-means algorithm is a partitioning clustering method that requires the number of clusters to be determined in advance. The algorithm classifies the data by repeated iterations until the stopping condition is met. Set the number of clusters k = 4, randomly select 4 initial centroids, calculate the distance from each data point to the 4 centroids, and assign each point to the nearest centroid to form 4 clusters. Recalculate the centroid of each cluster, that is, calculate the average value of all points in each cluster as the new centroid, check whether the change of the centroid is less than a certain threshold or reaches the maximum number of iterations. If not, repeat the assignment and update steps until the centroid position of each cluster no longer moves. According to the clustering threshold, the spatiotemporal risk index is divided into 4 safety risk levels: low risk, medium risk, higher risk and high risk, as shown in Table 5.

[0099] Table 5 Risk level classification results based on K-means clustering

[0100] Risk Level Low risk Medium risk Higher risk High risk TSRI [0，2.3467) [2.3467，2.6546) [2.6546，2.9224) [2.9224，+∞)

[0101] In order to further describe the technical solution of the present invention in detail, the section k103+100 to k113+100 of the Hebei section of Beijing-Harbin Expressway is taken as an example to calculate the spatiotemporal risk index of the section, analyze the distribution characteristics of the spatiotemporal risk index from the perspectives of space and time, and evaluate the traffic safety level of the section. Figure 3 As shown in the figure, most of the road sections have a low risk level. Figures 4 to 6 are the distribution of spatiotemporal risk index in different months, weeks and times, Figure 4 It can be seen that the time-space risk index is higher in April, July, August and October. Figure 5It can be seen that the time-space risk index is low from Monday to Thursday, and it is significantly higher from Friday to Sunday. Figure 6 It can be seen that the spatial-temporal risk index is higher in the morning, afternoon and night, while it is lower in the early morning and noon. The above results are convenient for highway traffic management departments to evaluate the road traffic safety status in real time and investigate the high-incidence period and accident-prone sections.

[0102] An embodiment of the present invention is described in detail above, but the described content is only a preferred embodiment of the present invention, and the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present invention, which should be covered within the protection scope of the present invention.

Claims

1. A method for evaluating highway traffic safety status, characterized in that: The method comprises the following steps: S1: Obtain historical highway accident data and road alignment data from relevant departments, match the accident data with the road alignment data, and obtain processed data; S2: Count the accident rates under various time indicators and road linear indicators, and formulate scoring standards for spatiotemporal risk indicators based on the accident rates; S3: The subjective weight of the time index is calculated by using the analytic hierarchy process, and the objective weight of the road alignment index is calculated by using the entropy weight method; S4: The spatiotemporal risk index model is constructed by integrating the risk scores and weight coefficients of each indicator and calculating the spatiotemporal risk index of the road; S5: Cluster the spatiotemporal risk index based on the K-means clustering algorithm and divide it into four security risk levels.

2. The method according to claim 1, characterized in that The acquisition of highway historical accident data and road alignment data includes: Obtain accident data, including: date, day of the week, time and pile number of the accident; Obtain road alignment data, including road plane alignment data and longitudinal section alignment data.

3. The method according to claim 1, characterized in that The matching of the accident data with the road alignment data includes: Eliminate data with incomplete information records, and obtain the horizontal and vertical alignment indicators of the road at the accident site through the road alignment data based on the pile number where the accident occurred.

4. The method according to claim 1, characterized in that: The time-space risk index scoring standard based on the accident rate includes: An accident rate of 0-10% is low risk, with a score of 1; an accident rate of 10-15% is medium risk, with a score of 2; an accident rate of 15-20% is relatively high risk, with a score of 3; an accident rate of more than 20% is high risk, with a score of 4.

5. The method according to claim 1, characterized in that The method of using the analytic hierarchy process to calculate the subjective weight of the time index and the method of using the entropy weight method to calculate the objective weight of the road alignment index includes: The analytic hierarchy process is used to calculate the weight of each time indicator, which includes: month, week, and time; The entropy weight method is used to calculate the weight of each road alignment index, which includes: straight line length, circular curve radius, curve deflection angle, uphill slope, downhill slope, slope length, and vertical curve radius.

6. The method according to claim 1, characterized in that The spatiotemporal risk index model is constructed by integrating the risk scores and weight coefficients of various indicators, including: The calculation formula of the Time-Space Risk Index (TSRI) is as follows: TSRI=w TRI TRI+w ARI DAY Where: TSRI is the spatiotemporal risk index; TRI is the time risk index; ARI is the linear risk index; w k is the weight of the risk item in TSRI, k = {TRI, ARI}; The TRI calculation formula is as follows: <h2 style=";text-align:left;direction:ltr">TRI = [T1 T2 T3]×[S1 S2 S3]<h2 style=";text-align:left;direction:ltr"> T Where: T i is the risk level under the i-th time index; S i is the weight of the i-th time indicator; The ARI calculation formula is as follows: ARI=[X1 X2 X3 X4 X5 X6 X7]× [ω1 ω2 ω3 ω4 ω5 ω6 ω7] T Where: X i is the risk level of the th linear indicator; ω i is the weight of the i-th linear indicator; The weight of each risk item in the spatiotemporal risk index is calculated as follows: Where: k is the sample standard deviation of the time risk index or linear risk index; k = {TRI, ARI}.

7. The method according to claim 1, characterized in that The K-means clustering algorithm is used to cluster the spatiotemporal risk index and divide it into four security risk levels, including: Set the number of clusters k = 4, randomly select 4 initial centroids, calculate the distance from each data point to the 4 centroids, and assign each point to the nearest centroid to form 4 clusters. Recalculate the centroid of each cluster, that is, calculate the average of all points in each cluster as the new centroid, and check whether the change of the centroid is less than a certain threshold or reaches the maximum number of iterations. If not, repeat the assignment and update steps until the centroid position of each cluster no longer moves. According to the clustering threshold, the spatiotemporal risk index is divided into 4 safety risk levels: low risk, medium risk, higher risk and high risk.

Images