A method for measuring highway traffic congestion based on radar trajectory data

Abstract

This invention discloses a method for measuring highway traffic congestion based on radar trajectory data. First, the invention obtains characteristics such as the speed and lane position of each vehicle from radar trajectory data, calculates the vehicle speed variation coefficient and lane change frequency, and calculates the traffic flow disturbance perception index per minute based on the CRITIC weighting method to represent the unstable state of traffic flow. Then, it calculates the traffic operation efficiency by collecting the traffic flow per minute. Finally, it combines the traffic flow disturbance perception index and traffic operation efficiency to calculate the traffic congestion degree, and determines the traffic congestion level based on the ROC curve. This invention's measurement of congestion is not limited to macroscopic traffic flow parameters; by scientifically setting traffic congestion level thresholds, it can more accurately reflect the congestion degree of highway mainline sections or weaving sections.

Description

Technical Field

[0001] This invention relates to a method for measuring highway traffic congestion based on radar trajectory data, belonging to the field of traffic congestion evaluation technology. Background Technology

[0002] Existing methods for measuring traffic congestion mostly rely on manually set congestion thresholds, which do not comprehensively consider the characteristics of traffic congestion and are highly subjective in their congestion assessment. Summary of the Invention

[0003] The purpose of this invention is to extract traffic flow instability and traffic operation efficiency indicators based on radar trajectory data, reflecting the degree of traffic congestion.

[0004] To achieve the above objectives, the technical solution of the present invention provides a method for measuring highway traffic congestion based on radar trajectory data, characterized by comprising the following steps:

[0005] Step S-1: Acquire millimeter-wave radar trajectory data;

[0006] Step S-2: After extracting the ID, time t, speed V(t), and lane l of each vehicle, calculate the speed variation coefficient C of the vehicle at time t. v () and the lane change frequency R(t):

[0007]

[0008] In the formula: N is the number of lanes; s t,l Let be the standard deviation of the vehicle speeds in lane l at time t; Let be the average speed of vehicles in lane l at time t;

[0009]

[0010] Where: n l (t) represents the number of lane changes in lane l at time t; L represents the length of the road segment;

[0011] Step S-3: Calculate the traffic flow disturbance perception index s from time t to time t+60:

[0012]

[0013] In the formula: s(t) is the traffic flow disturbance perception index at time t; ω1 and ω2 are the vehicle speed variation coefficient and the weight of lane change frequency;

[0014] Step S-4: Obtain the traffic flow q from time t to time t+60 collected by the radar;

[0015] Step S-5: Calculate the running efficiency E:

[0016]

[0017] Step S-6: Calculate traffic congestion level D:

[0018]

[0019]

[0020] In the formula, E e To achieve optimal operating efficiency, a curve showing the operating efficiency versus traffic flow per minute is plotted based on historical data; the extreme point of this curve represents the optimal operating efficiency. min D represents the minimum traffic congestion level. max This indicates the maximum level of traffic congestion.

[0021] Preferably, in step S-2, the number of lane changes n of lane l at time t is... l The calculation method for (t) includes the following steps:

[0022] Determine whether vehicles with the same ID are in the same lane (l) at time t-1 as they are at time t. If so, determine the number of lane changes (n) in lane l at time t. l (t)= l (t); if not, then the number of lane changes n at time t. l (t)= l (t)+1.

[0023] Preferably, in step S-3, the weights ω1 and ω2 are based on the CRITIC weighting method, which comprehensively measures the objective weights by comparing the intensity and conflict between the speed variation coefficient and the lane change frequency. The calculation formula is as follows:

[0024]

[0025] In the formula, s i γ is the speed variation coefficient and the standard deviation of lane change frequency calculated based on historical data; γ is the correlation coefficient of speed variation coefficient and lane change frequency calculated based on historical data.

[0026] Preferably, in step S-6, the congestion level is also determined based on the ROC curve.

[0027] Preferably, determining the congestion level based on the ROC curve includes the following steps:

[0028] Step S-6.1: Calculate the average travel speed from historical data.

[0029] Step S-6.2: According to the "Road Traffic Congestion Evaluation Method" GA / T 115-2020, obtain samples of four congestion levels: smooth traffic, light congestion, moderate congestion, and heavy congestion.

[0030] Step S-6.3: Plot ROC curves for each of the following two categories: smooth traffic and light congestion, light congestion and moderate congestion, and moderate congestion and severe congestion.

[0031] Step S-6.4: Select the sample thresholds D3, D2, and D1 that are closest to the (0,1) point on the ROC curves of smooth and lightly congested, lightly congested and moderately congested, and moderately congested and severely congested as the optimal thresholds;

[0032] Step S-6.5: Obtain the area under the ROC curve (AUC);

[0033] Step S-6.6: Determine whether the AUC of adjacent congestion level thresholds is greater than or equal to 0.5, where 0.5 is the standard for judging the validity of ROC curve classification in mathematical analysis.

[0034] Step S-6.6.1: If it is less than 0.5, the sample is reclassified according to the congestion level threshold and returned to step S-6.2;

[0035] Step S-6.6.2: If the value is greater than or equal to 0.5, determine whether the AUC value is the same as the previous round: if yes, output the optimal thresholds D3, D2, and D1 when the AUC is at its maximum, and formulate a traffic congestion level correspondence table; if no, return to step S-6.6.1.

[0036] This invention considers traffic congestion not only from macroscopic traffic flow parameters, but also combines traffic flow instability and traffic operation efficiency to propose a novel method for measuring highway traffic congestion based on radar trajectory data. It uses ROC curves to set traffic congestion level thresholds, avoiding subjectivity caused by manual settings. This method meets standard requirements and can more accurately reflect the congestion level of highway main road sections or weaving sections, thus solving the shortcomings of existing technologies. Attached Figure Description

[0037] Figure 1 A schematic diagram of a highway traffic congestion measurement method based on radar trajectory data;

[0038] Figure 2 This is a flowchart illustrating the process of determining traffic congestion levels based on ROC curves. Detailed Implementation

[0039] The present invention will be further illustrated below with reference to specific embodiments. It should be understood that these embodiments are for illustrative purposes only and are not intended to limit the scope of the invention. Furthermore, it should be understood that after reading the teachings of this invention, those skilled in the art can make various alterations or modifications to the invention, and these equivalent forms also fall within the scope defined by the appended claims.

[0040] Combination Figure 1 and Figure 2 The present invention provides a method for measuring highway traffic congestion based on radar trajectory data, comprising the following steps:

[0041] Step S-1: Acquire millimeter-wave radar trajectory data with a time accuracy of seconds.

[0042] Step S-2: Extract the ID, time t, speed V(t), and lane l of each vehicle, further including the following steps:

[0043] Step S-2.1.1: Calculate the standard deviation and average speed of vehicles in each lane at time t, where the standard deviation of vehicle speed in lane l is expressed as s. t,l Average speed is expressed as

[0044] Step S-2.1.2: Calculate the velocity variation coefficient C of the vehicle at time t. v ():

[0045] The coefficient of variation of speed reflects the speed instability of a convoy following another vehicle in the same lane. To avoid the influence of the mean on the dispersion, a combination of the speed standard deviation and the mean speed is used for calculation, as shown in the following formula:

[0046]

[0047] In the formula, N is the number of lanes;

[0048] Step S-2.2.1: Determine whether cars with the same ID are in the same lane at time t-1 and time t:

[0049] If so, then the number of lane changes n in lane l at time t. l (t)= l (t); if not, then the number of lane changes n at time t. l (t)= l (t)+1;

[0050] Step S-2.2.3: Calculate the vehicle lane-changing frequency R(t) at time t:

[0051] The frequency of lane changes within a road segment is quantified by lane change frequency, reflecting the degree of mutual interference between traffic flows in different lanes. The calculation formula is as follows:

[0052]

[0053] In the formula, L is the length of the road segment.

[0054] Step S-3: Calculate the traffic flow disturbance perception index s from time t to time t+60:

[0055] The Traffic Flow Disturbance Perception Index comprehensively reflects the traffic flow instability caused by vehicle speed instability and mutual interference. The calculation formula is as follows:

[0056]

[0057] In the formula, S(t) is the traffic flow disturbance perception index at time t; ω1 and ω2 are the vehicle speed variation coefficients and the weights of lane change frequency.

[0058] Weights ω1 and ω2 are based on the CRITIC weighting method, which comprehensively measures the objective weights by comparing the intensity and conflict between the speed variation coefficient and the lane change frequency. The calculation formula is as follows:

[0059]

[0060] In the formula, s i γ is the speed variation coefficient and the standard deviation of lane change frequency calculated based on historical data; γ is the correlation coefficient of speed variation coefficient and lane change frequency calculated based on historical data.

[0061] Step S-4: Obtain the traffic flow q from time t to time t+60 collected by the radar.

[0062] Step S-5: Calculate the operating efficiency E, as shown in the following formula:

[0063]

[0064] Step S-6: Calculate the traffic congestion degree D and determine the congestion level based on the ROC curve:

[0065] Traffic congestion level comprehensively reflects the instability of traffic flow and traffic operation efficiency. The calculation formula is as follows:

[0066]

[0067]

[0068] In the formula, E e To achieve optimal operating efficiency, a curve showing the operating efficiency versus traffic flow per minute is plotted based on historical data; the extreme point of this curve represents the optimal operating efficiency. min D represents the minimum traffic congestion level. max This indicates the maximum level of traffic congestion.

[0069] Determining congestion levels based on ROC curves includes the following steps:

[0070] Step S-6.1: Calculate the average travel speed from historical data.

[0071] Step S-6.2: According to the "Road Traffic Congestion Evaluation Method" GA / T 115-2020, as shown in Table 1, obtain samples of four congestion levels: smooth traffic, light congestion, moderate congestion, and heavy congestion.

[0072] Table 1. Evaluation Method for Road Traffic Congestion (GA / T 115-2020)

[0073]

[0074] Step S-6.3: Plot ROC curves for each of the following two categories: smooth traffic and light congestion, light congestion and moderate congestion, and moderate congestion and severe congestion.

[0075] Step S-6.4: Select the sample thresholds D3, D2, and D1 that are closest to the (0,1) point on the ROC curves of smooth and lightly congested, lightly congested and moderately congested, and moderately congested and severely congested as the optimal thresholds;

[0076] Step S-6.5: Obtain the area under the ROC curve (AUC);

[0077] Step S-6.6: Determine whether the AUC of adjacent congestion level thresholds is greater than or equal to 0.5, where 0.5 is the standard for judging the validity of ROC curve classification in mathematical analysis.

[0078] Step S-6.6.1: If it is less than 0.5, the sample is reclassified according to the congestion level threshold and returned to step S-6.2;

[0079] Step S-6.6.2: If the value is greater than or equal to 0.5, determine whether the AUC value is the same as the previous round. If so, output the optimal thresholds D3, D2, and D1 when the AUC is at its maximum, and formulate a traffic congestion level correspondence table. The values ​​are obtained using this method based on the provisions of "Road Traffic Congestion Evaluation Method" GA / T 115-2020, as shown in Table 2. If not, return to step S-6.6.1.

[0080] Table 2. Correspondence between Traffic Congestion Levels

[0081]

Claims

1. A method for measuring highway traffic congestion based on radar trajectory data, characterized in that, Includes the following steps: Step S-1: Acquire millimeter-wave radar trajectory data; Step S-2: After extracting the ID, time t, speed V(t), and lane l of each vehicle, calculate the speed variation coefficient C of the vehicle at time t. v (t) and the lane change frequency R(t): In the formula: N is the number of lanes; s t,l Let be the standard deviation of the vehicle speeds in lane l at time t; Let be the average speed of vehicles in lane l at time t; In the formula, n l (t) represents the number of lane changes in lane l at time t, where L is the road segment length, and n represents the number of lane changes in lane l at time t. l The calculation method for (t) includes the following steps: Determine whether vehicles with the same ID are in the same lane (l) at time t-1 as they are at time t. If so, determine the number of lane changes (n) in lane l at time t. l (t)=n l (t); if not, then the number of lane changes n at time t. l (t)=n l (t)+1; Step S-3: Calculate the traffic flow disturbance perception index S from time t to time t+60. In the formula: S(t) is the traffic flow disturbance perception index at time t; ω1 and ω2 are the vehicle speed variation coefficient and the weight of lane change frequency; Step S-4: Obtain the traffic flow q from time t to time t+60 collected by radar; Step S-5: Calculate the running efficiency E: Step S-6: Calculate traffic congestion level D: In the formula, E e To achieve optimal operating efficiency, a curve showing the operating efficiency versus traffic flow per minute is plotted based on historical data; the extreme point of this curve represents the optimal operating efficiency. min D represents the minimum traffic congestion level. max This indicates the maximum level of traffic congestion.

2. The method for measuring highway traffic congestion based on radar trajectory data as described in claim 1, characterized in that, In step S-3, the weights ω1 and ω2 are based on the CRITIC weighting method, which comprehensively measures the objective weights by comparing the intensity and conflict between the speed variation coefficient and the lane change frequency. The calculation formula is as follows: In the formula, s i γ is the speed variation coefficient and the standard deviation of lane change frequency calculated based on historical data; γ is the correlation coefficient of speed variation coefficient and lane change frequency calculated based on historical data.

3. The method for measuring highway traffic congestion based on radar trajectory data as described in claim 1, characterized in that, In step S-6, the congestion level is also determined based on the ROC curve.

4. The method for measuring highway traffic congestion based on radar trajectory data as described in claim 1, characterized in that, Determining congestion levels based on ROC curves includes the following steps: Step S-6.1: Calculate the average travel speed from historical data. Step S-6.2: According to the "Road Traffic Congestion Evaluation Method" GA / T 115-2020, obtain samples of four congestion levels: smooth traffic, light congestion, moderate congestion, and heavy congestion. Step S-6.3: Plot ROC curves for each of the following two categories: smooth traffic and light congestion, light congestion and moderate congestion, and moderate congestion and severe congestion. Step S-6.4: Select the sample thresholds D3, D2, and D1 that are closest to the (0,1) point on the ROC curves of smooth and light congestion, light congestion and moderate congestion, and moderate congestion and severe congestion as the optimal thresholds; Step S-6.5: Obtain the area under the ROC curve (AUC); Step S-6.6: Determine whether the AUC of adjacent congestion level thresholds is greater than or equal to 0.5, where 0.5 is the standard for judging the validity of ROC curve classification in mathematical analysis. Step S-6.6.1: If it is less than 0.5, the sample is reclassified according to the congestion level threshold and returned to step S-6.2; Step S-6.6.2: If the value is greater than or equal to 0.5, determine whether the AUC value is the same as the previous round: if yes, output the optimal thresholds D3, D2, and D1 when the AUC is at its maximum, and formulate a traffic congestion level correspondence table; if no, return to step S-6.6.1.

Images