Urban road congestion evaluation method considering macro-micro data coupling

Abstract

The application provides a kind of urban road congestion evaluation method considering macro-micro data coupling, comprising the following steps: S1: obtaining GIS geographic information data, modeling traffic network using complex network theory;S2: preprocessing vehicle GPS trajectory data, extracting speed information of each road section, obtaining traffic state index;S3: obtaining percolation threshold according to the speed information of each road section;S4: obtaining the running state threshold of each road section according to the percolation threshold;S5: obtaining the running state index according to the traffic state index, obtaining the system congestion index SCI in combination with the running state threshold.The application provides a road congestion evaluation method coupling macro and micro, aiming at the problem of urban road congestion evaluation, avoids the type and quality difference of detector brought by introducing multi-source data, cost benefit and the uncertainty of data real-time, and solves the problem of losing detailed information of specific road section caused by describing traffic condition from macro perspective only.

Description

Technical Field

[0001] This invention relates to the field of traffic network assessment technology, and in particular to a method for evaluating urban road congestion that considers macro-micro data coupling. Background Technology

[0002] With socio-economic development and accelerated urbanization, the contradictions arising from the limited urban transportation resources are becoming increasingly prominent, manifested in the imbalance between supply and demand in urban road networks, environmental pollution, and traffic safety issues. To address these complex traffic problems, a series of traffic control and management solutions are needed. Urban traffic conditions reflect the time-varying, multi-scale, multi-variable, and stochastic characteristics of traffic congestion in urban road networks. Quantifying the degree of traffic congestion allows transportation management departments to grasp the spatiotemporal information of congested road sections, serving as the foundation for implementing traffic control and management. Traffic flow in urban road networks is complex and dynamic, making a comprehensive evaluation of the congestion level of each road segment a challenging task. Therefore, it is urgent to establish a scientific method for evaluating the traffic conditions of urban road segments to provide a reference for efficient traffic management decisions.

[0003] Chinese invention patent, publication number CN106781491B, entitled "An Urban Road Traffic Condition Assessment System," discloses an urban road traffic condition assessment system comprising a data acquisition unit, a data processing unit, a data transmission unit, and a traffic condition assessment unit. The data acquisition unit collects traffic parameters at key locations; the data processing unit processes the collected data and transmits it to the traffic condition assessment unit via the data transmission unit; and the traffic condition assessment unit assesses the road traffic condition based on the processed traffic parameter values. This technical solution comprehensively considers vehicle status information and road traffic status information, establishing a comprehensive and efficient road traffic condition assessment system, achieving effective assessment of road traffic conditions. However, this technical solution has strict requirements for road sensor placement and cannot perform assessments on road sections lacking sensor data support.

[0004] Chinese invention patent, publication number CN108777064A, entitled "A Traffic State Assessment System Based on Information Fusion," discloses a traffic state assessment system based on information fusion. This system includes a traffic information detection module, a data transmission module, a data preprocessing module, a feature extraction module, a state assessment module, and a result evaluation module. The traffic information detection module collects traffic information; the data transmission module transmits the traffic information to the data preprocessing module, which filters and removes duplicates from the traffic information; the feature extraction module extracts and normalizes features from the processed traffic information; the state assessment module evaluates the traffic state; and the result evaluation module calculates the accuracy of the traffic state assessment results and issues a warning when the accuracy is low. This technical solution uses evidence theory to fuse traffic monitoring information, overcoming its shortcomings in handling conflicting evidence through improvements to evidence theory. However, the information fusion process in this technical solution is coarse and lacks a single dimension, leading to inaccurate traffic state assessment results.

[0005] Chinese invention patent, publication number CN116030631A, entitled "A Real-Time Traffic Congestion Assessment Method Based on UAV Aerial Video," discloses a method for assessing real-time traffic congestion based on UAV aerial video. This method consists of a traffic flow statistics module and a road time occupancy calculation module. The traffic flow statistics module records the tracked vehicle IDs and detection box coordinates. When the change in the detection box coordinates of a vehicle with the same ID exceeds a threshold, the traffic flow is incremented by 1. The road time occupancy calculation module sets up N dynamically changing virtual coils and records the number of times each virtual coil is occupied by a vehicle in each frame of the image, calculating the road time occupancy rate. Traffic congestion is assessed by combining traffic flow and road time occupancy within a specified time period. This technical solution solves the problems of existing methods requiring significant manpower and resources, equipment aging and damage, and complex implementation. It balances accuracy, real-time performance, and economy, and is of great significance for traffic congestion assessment and traffic scheduling management. However, this technical solution has a limited detection range and uses a single congestion assessment indicator, which can easily lead to a mismatch between the assessment results and the actual situation. Summary of the Invention

[0006] To address the problems in existing technologies where strict road sensor placement schemes and single congestion assessment indicators lead to inaccurate or impossible congestion assessments, this invention proposes a city road congestion evaluation method that considers macro-micro data coupling. This method couples macro and micro levels of road congestion assessment, avoiding uncertainties related to detector type and quality differences, cost-effectiveness, and data real-time performance arising from multi-source data. It also solves the problem of losing detailed information about specific road segments due to describing traffic conditions only from a macro perspective.

[0007] This invention is achieved through the following technical solution, including the following steps:

[0008] S1: Obtain GIS geographic information data within the evaluation road network area, and model the traffic network within the evaluation road network area using complex network theory;

[0009] S2: Preprocess the vehicle GPS trajectory data within the evaluation road network area to extract the speed information of each road segment. The speed information of each road segment includes the current speed v of each road segment within the time interval t. ij (t) and the maximum speed limit Obtain the traffic status index U(T) of the road network area within the time interval T;

[0010] S3: Based on the current speed v of each road segment within time interval t. ij (t) and the maximum speed limit The infiltration threshold q within the evaluation road network area is obtained. c (T);

[0011] S4: Based on the seepage threshold q c (T), to obtain the operating status threshold θ(T) for each road segment;

[0012] S5: Based on the traffic status index U(t) of the evaluated road network area within the time interval T, the operational status index P(t) of the evaluated road network area is obtained. Combined with the operational status threshold θ(T), the system congestion index SCI is obtained.

[0013] As a further preferred option, the specific steps of step S1 are as follows:

[0014] S1-1. Model the traffic network as a graph G(R, E, W), where R = {r1, ..., r2} n} is a set of nodes, where n is the number of nodes, and E = {e} ij Let {i,j=1,...,k} be the set of road segments, and e ij To connect node r i and node r j The road segments, k is the number of road segments, W = {wij The time weight matrix is ​​defined by {(t), i, j = 1, ..., k}, where w ij (t) represents road segment e ij The running status within time interval t.

[0015] As a further preferred option, the specific steps of step S2 are as follows:

[0016] S2-1. Obtain vehicle GPS trajectory data within the evaluation road network area, perform preprocessing, and obtain speed information for each road segment, i.e., the current speed v of each road segment. ij (t) and the maximum speed limit within the time interval t The maximum speed limit within the time interval t The 95th percentile of the current velocity within time interval t;

[0017] S2-2. Based on the speed information of each road segment, the traffic state index U(T) within the time interval T is obtained using formula (1).

[0018] U(T)=g(w ij (t)), t=t m T = t M (1)

[0019] In the formula, Time interval t M Greater than the time interval t m .

[0020] As a further preferred option, the time interval t M The interval is 60 minutes, and the time interval is t. m It lasts for 5 minutes.

[0021] As a further preferred option, the specific steps of step S3 are as follows:

[0022] S3-1. Based on the current speed v of the road segment within time interval t. ij (t), time interval t M and time interval t m The e of each road segment within time period T can be obtained from formula (2). ij average velocity v ij (T):

[0023]

[0024] S3-2. Based on the time period T, each road segment e ij average velocity v ij (T) and maximum speed limit The functional state f of each road segment is obtained from formula (3). ij(T):

[0025]

[0026] Where q(T) is the threshold value of the road segment during time interval T;

[0027] S3-3. Select f from step S3-2 ij The penetration threshold q for evaluating the road network area is obtained when the second largest connected cluster in the connected cluster formed by road segments with T = 1 reaches its maximum value. c (T).

[0028] As a further preferred option, the value of q(T) is in the range of 0 to 1, where 0 indicates that the road segment is completely blocked and all vehicles are stationary, and 1 indicates that the road segment is free to flow and all vehicles are traveling at maximum speed.

[0029] As a further preferred option, the specific steps of step S4 are as follows:

[0030] S4-1. Based on the seepage threshold q c (T), the infiltration threshold q for evaluating the road network area is obtained from formula (4). c Change in (T)

[0031]

[0032] in, To pass through closed road section e ij And measure the updated penetration threshold of the modified evaluation road network area;

[0033] S4-2. Based on the seepage threshold q c (T) and the permeation threshold q c Change in (T) The operational status threshold θ(T) of the evaluated road network area is obtained;

[0034] θ(T) is determined by the e values ​​of each road segment in the evaluation road network area. ij The dynamic threshold θ of the running state ij (T) composition, θ ij (T) is obtained from formula (5).

[0035]

[0036] Where λ is the correction coefficient.

[0037] As a further preferred option, the specific steps of step S5 are as follows:

[0038] S5-1. Based on the traffic state index U(T) of the road network within the time interval T obtained in step S2-2, obtain the operational state index P(t) of the road network area, where P(t) = U(T);

[0039] S5-2. Based on the operational status index P(t) of each road segment within the evaluation road network area from 0:00 to 6:00 AM, the free flow range of the evaluation road network area is obtained. The minimum value of the free flow range is the lower bound index P(t) of the free flow range. pre );

[0040] S5-3. Based on the operating state index P(t) and the operating state threshold θ(T), the resistance coefficient R is obtained from formula (6). e (t),

[0041]

[0042] S5-4. Based on the operating status index P(t) and the lower bound index P(t) of the free flow range. pre The congestion level C can be obtained from formula (7). m (t),

[0043]

[0044] S5-5. Based on the drag coefficient R e (t) and congestion level C m (t), from the formula SCI(t) = C m (t)+R e (t) yields the system congestion index SCI.

[0045] As a further preferred option, the preprocessing method for vehicle GPS trajectory data in step S2-1 is as follows: the vehicle GPS trajectory data is matched with the road network of the evaluation road network area using a map matching algorithm, and the dense GPS data accumulated at the locations of stationary vehicles or other vehicles that have been parked for a long time is removed using the DBSCAN clustering method.

[0046] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0047] 1. The method of the present invention evaluates road congestion by generating a systematic congestion index through macro-micro coupling, which bridges the gap between evaluation of a single road segment and macro-level evaluation of the road network, and can provide a more accurate assessment of the road network traffic status, which helps to implement effective traffic management and control measures on specific road segments.

[0048] 2. The method of the present invention uses the seepage theory to describe the macroscopic traffic state of the road network and uses this as an evaluation baseline for correction, which can obtain a reliable congestion assessment without introducing additional data sources.

[0049] 3. This invention avoids the problems of detector type and quality differences, cost-effectiveness and inaccurate data real-time performance caused by introducing multiple data sources (such as inductive loops, traffic monitoring videos, and radio frequency identification). The method of obtaining evaluation data sources is simple and efficient.

[0050] 4. This invention uses the DBSCAN clustering method for data preprocessing, which can effectively remove noisy data and improve the reliability of evaluation results. Attached Figure Description

[0051] Figure 1 This is a flowchart of the evaluation method in this invention.

[0052] Figure 2 This invention provides a framework for assessing road congestion status.

[0053] Figure 3(a) shows the distribution of GPS data.

[0054] Figure 3(b) shows the speed distribution of the road segment after GIS preprocessing.

[0055] Figure 4(a) is a schematic diagram of the microwave detector arrangement.

[0056] Figure 4(b) is a schematic diagram of the congestion assessment results using SCI as an indicator.

[0057] Figure 4(c) is a schematic diagram of the congestion assessment results using occupancy rate as an indicator.

[0058] Figure 4(d) is a schematic diagram of the congestion assessment results using relative speed as an indicator.

[0059] Figure 5 This is a schematic diagram illustrating the assessment of road network congestion using SCI as an indicator. Detailed Implementation

[0060] The advantages and features of the present invention will be illustrated and explained by the following non-limiting description of preferred embodiments, which are given by way of example only with reference to the accompanying drawings.

[0061] like Figure 1 As shown, this invention proposes a method for evaluating urban road congestion that considers macro-micro data coupling. This method includes the following steps:

[0062] S1: Obtain GIS geographic information data within the evaluation road network area, and use complex network theory to model the traffic network within the evaluation road network area.

[0063] S1-1. Obtain GIS geographic information data within the evaluation road network area and model the traffic network within the evaluation road network area. Model the traffic network as a graph G(R, E, W), where R = {r1, ..., r...}n} is the set of nodes (i.e., road intersections), where n is the number of nodes, and E = {e ij Let {i,j=1,...,k} be the set of road segments, and e ij To connect node r i and node r j The road segments, k is the number of road segments, W = {w ij {w(t), i, j = 1, ..., k} is a series of time weight matrices, where w ij (t) represents road segment e ij The running status within time interval t.

[0064] S2: Preprocess the vehicle GPS trajectory data within the evaluation road network area, extract the speed information of each road segment, and obtain the traffic status index (traffic congestion level) U(T) of the evaluation road network area within the time interval T.

[0065] S2-1. Obtain vehicle GPS trajectory data within the evaluation road network area, perform preprocessing, and obtain speed information for each road segment.

[0066] GPS trajectory data of vehicles within the evaluation road network area was acquired, and speed information of each road segment was extracted. The acquired GPS data was matched with the road network of the evaluation road network area using a map matching algorithm. Dense GPS data accumulated at locations where stationary vehicles or other vehicles had been parked for extended periods was removed using the DBSCAN clustering method; for road segments where speed data could not be collected during certain time periods, interpolation methods were used to fill in the missing data; for roads without any speed data, the average speed of connecting roads was used as a substitute. This process was repeated at time intervals t until the time interval reached T, obtaining GPS data for all road segments within time interval T, which was then plotted as shown in Figure 3(a). Using a 5-minute time interval t, the current speed v of each road segment was obtained using GPS data. ij (t), plotted as a graph, as shown in Figure 3(b), and the maximum speed limit within the time interval t. The maximum speed limit within the time interval t The 95th percentile of the current speed within time interval t. The current speed v of each road segment within time interval t. ij (t) and the maximum speed limit This provides speed information for each road segment.

[0067] S2-2. Based on the speed information of each road segment, the traffic state index (traffic congestion level) U(T) of the road network area within the time interval T is obtained and determined by formula (1):

[0068] U(T)=g(w ij (t)), t=tm T = t M (1)

[0069]

[0070] In equation (1), based on the spatiotemporal mapping function g(·) and road segment e ij The running state w within time interval t ij (t) yields the traffic state index U(T) within the evaluated road network area over time interval T. Where, time interval t M Greater than the time interval t m The specific time interval t M The interval is 60 minutes, and the time interval is t. m It lasts for 5 minutes.

[0071] In equation (2), w ij (t) represents the current velocity v ij (t) and its maximum speed limit within time interval t The ratio of v to v. ij (t) represents road segment e ij The current speed within time interval t It is the 95th percentile of its historical speed.

[0072] S3: Based on the seepage theory, analyze the congestion diffusion effect of the road network and obtain the seepage threshold q in the evaluation area of ​​the road network. c (T) serves as traffic status information at the macro-level road network.

[0073] S3-1. Based on the current speed v of the road segment within time interval t. ij (t), time interval t M and time interval t m Get the e of each road segment within time period T ij average velocity v ij (T), is determined by formula (3):

[0074]

[0075] S3-2. Based on the time period T, each road segment e ij average velocity v ij (T) and maximum speed limit Obtain the functional state f of each road segment ij (T), is determined by formula (4):

[0076]

[0077] Where q(T) is the threshold for the road segment during time interval T. Based on the considered acceptable speed level, q(T) is used to determine whether the road segment reaches maximum functionality within time interval T. The value of q(T) ranges from 0 to 1, where 0 represents complete road blockage with all vehicles stationary, and 1 represents free-flowing road with all vehicles traveling at maximum speed. The average operating state recorded within time interval T is also considered. Greater than road segment e ij If the operating status recorded by q(T) is used, then the road segment is considered a functionally normal road segment (i.e., f). ij (T) = 1), otherwise, it is a dysfunctional segment (i.e., f). ij (T) = 0).

[0078] S3-3. Select f from step S3-2 ij The penetration threshold q for evaluating the road network area is obtained when the second largest connected cluster in the connected cluster formed by road segments with T = 1 reaches its maximum value. c (T).

[0079] Using seepage theory, record f in step S3-2. ij The segments with q(T) = 1 form connected clusters. After removing dysfunctional segments from the network, each segment will form different biconnected segments within a given time interval T. The value of q(T) when the second largest connected cluster reaches its maximum value is defined as the penetration threshold q of the road network. c (T). This reflects the optimal state that the evaluated road network area can achieve within time period T, where vehicles can travel smoothly at relative speeds on most road sections. q c (T) represents the phase transition point of functional transportation network connectivity, which can effectively measure the maximum operating state and reflect its global efficiency.

[0080] S4: Based on the seepage threshold q c (T), to obtain the operation state threshold (OST)θ(T) for each road segment.

[0081] S4-1. Based on the seepage threshold q c (T) Obtain the penetration threshold q for the evaluated road network area. c Change in (T) Determined by formula (5):

[0082]

[0083] in, To pass through the closed section (segment e) ij The running status is changed to 0) and the updated penetration threshold of the modified evaluation road network area is measured.

[0084] S4-2. Based on the seepage threshold q c (T) and the permeation threshold q c Change in (T) The Operation State Threshold (OST)θ(T) of the evaluated road network area is obtained.

[0085] θ(t) is determined by the e values ​​of each road segment in the evaluation road network area. ij The dynamic threshold θ of the running state ij (T) composition. Constructed θ ij (T) can be expressed by formula (6):

[0086]

[0087] Where λ is the correction factor. Note that if road segment e ij The running status increases from 0 to 1 and exceeds If the value is high, then the road segment tends to improve the efficiency of the road network and achieve better macro-traffic conditions (i.e., alleviate traffic congestion).

[0088] S5: Based on the operating state threshold θ(T), the Systemic Congestion Index (SCI) is obtained to evaluate the congestion status of each road segment within the road network area.

[0089] S5-1. Based on the traffic state index U(T) of the road network within the time interval T obtained in step S2-2, the operational state index P(t) of the road network area is obtained, where P(t) = U(T). Figure 2 As shown.

[0090] S5-2. Based on the operational status index P(t) of each road segment within the evaluation road network area from 0:00 to 6:00 AM, the free flow range of the evaluation road network area is obtained. The minimum value of the free flow range is the lower bound index P(t) of the free flow range. pre ).

[0091] Traffic flow within the free flow range is considered non-congested. When the operational status index P(t) is lower than the lower bound index P(t) of the free flow range... pre When this happens, congestion will occur, such as Figure 2 As shown.

[0092] S5-3. Based on the operating state index P(t) and the operating state threshold θ(T), the drag coefficient R is obtained. e (t). This represents the severity of congestion spreading from a traffic state change on one road segment to adjacent road segments, and it is closely related to the operating state threshold θ(T). R eThe expression for (t) is shown in formula (7):

[0093]

[0094] If P(t) < θ(T), then the portion exceeding the lower bound is represented by R. e (t)=θ(T)-P(t). The critical threshold for congestion diffusion is (C t When P(t) < θ(T), this road segment will cause congestion to spread between adjacent road segments, thus accelerating the spread of congestion, such as... Figure 2 As shown in Figure C. t (t) can be expressed by formula (8):

[0095] C t (t)=P(t pre )-θ(T) (8)

[0096] S5-4. Based on the operating status index P(t) and the lower bound index P(t) of the free flow range. pre ), thus obtaining the congestion level C. m (t) is used to measure the degree of traffic congestion. It is determined by formula (9):

[0097]

[0098] When the road segment operation status index P(t) is at the lower bound of the free flow range, i.e., P(t) pre The decrease in operating state can be viewed as the process of congestion formation, where C > P(t). m (t)=P(t pre )-P(t). If P(t) pre )≤P(t), P(t) is still within the free flow range, where C m (t) = 0, as Figure 2 As shown.

[0099] S5-5. Based on the drag coefficient R e (t) and congestion level C m (t), thus obtaining the Systemic Congestion Index (SCI), which is used for evaluating the traffic conditions of urban road segments. It is expressed by formula (10):

[0100] SCI(t)=C m (t)+R e (t) (10)

[0101] During the changes in the operational status of a road segment, congestion forms and dissipates correspondingly on a macroscopic scale. Therefore, θ(T) is also affected by the macroscopic traffic conditions.

[0102] To further verify the feasibility and superiority of the urban road congestion evaluation method considering macro-micro data coupling described in this invention, a local road network in Kunshan, China, was selected. This area has 209 road segments and 144 intersections. Taxi GPS datasets were recorded every 5 seconds from January 4th to January 6th, 2018. Preprocessing was performed according to steps S1 to S2, resulting in Figures 3(a) and 3(b). Figure 3(a) shows the GPS data distribution of a local road network in Kunshan, China, from January 4th to January 6th, 2018. Figure 3(b) shows the spatial distribution of speeds at 12:00 on January 4th, 2018. The local road network data was then sequentially substituted into steps S3 to S5 to obtain the seepage threshold q for this local road network. c (T), operating status threshold θ(T) for each road segment, and resistance coefficient R e (t) and congestion level C m (t), ultimately yielding the System Congestion Index (SCI). Matching the SCI with the map of this local road network yields, as shown below. Figure 5 The diagram shown is an assessment of the road network congestion status. Figure 5 This is a road network congestion assessment map for three time periods: 6:30 AM, 7:30 AM, and 8:30 AM. It can assess the road congestion situation in this specific area of ​​Kunshan.

[0103] Visualization of road network congestion assessment using the Congestion Index (SCI) proposed in this invention, as shown below. Figure 5 As shown, around 6:30 AM, the SCI (Self-Congestion Index) identified most road sections as relatively smooth. However, during the morning rush hour (7:30 AM), congestion significantly worsened, with a marked increase in road sections within the 0.4-0.6 SCI range, and a further surge in the number of red-marked road sections, indicating severe congestion. One hour later (8:30 AM), congestion gradually eased, and most road sections reached a relatively stable state.

[0104] There are many indicators used to measure urban road traffic conditions. To verify the effectiveness of the System Congestion Index (SCI) proposed in this invention, relative speed (RV) and occupancy rate are introduced to assess the degree of congestion and compared with the trend of the SCI. Occupancy rate can be constructed based on the density and average speed of each road segment. Considering the high detection cost, currently only a few road segments are equipped with microwave detectors that can provide occupancy rate information.

[0105] As shown in Figure 4(a), the main roads selected for analysis include road segments 2, 12, 44, 62, and 76. Among them, road segments 12 and 76 are equipped with microwave detectors.

[0106] Figures 4(b), 4(c), and 4(d) show the congestion levels estimated by the System Congestion Index (SCI), occupancy rate, and relative speed (RV), respectively. The congestion level varies spatially and temporally. As shown in Figure 4(d), there is a relatively high level of congestion during the early morning hours. This is supported by the occupancy rate observed in Figure 4(c), which indicates that traffic demand during the early morning hours is typically lower than during the day. The situation described in Figure 4(d) can be explained by the unreliable and counterintuitive assessments caused by taxi trajectories at low GPS sampling rates during the early morning hours. The System Congestion Index (SCI) and the original occupancy rate assessment provided by this invention do not have such errors. Therefore, the System Congestion Index (SCI) provided by this invention is more accurate than the relative speed (RV) assessment.

[0107] As can be seen from Figures 4(b) and 4(c), higher levels of congestion can be observed during the morning rush hour (6:00 AM to 8:00 AM) and evening rush hour (7:00 PM to 9:00 PM) compared to other time periods. Furthermore, congestion variations are significant across different sections of this main road during the same time of day, while the variations shown in Figure 4(d) are relatively lower. Therefore, the System Congestion Index (SCI) provided by this invention is more accurate than the relative speed (RV) assessment; and compared to the need to install microwave detectors for occupancy assessment, the SCI provided by this invention saves more resources and costs.

[0108] In addition to the above embodiments, the present invention may have other implementation methods. All technical solutions formed by equivalent substitution or equivalent transformation fall within the protection scope claimed by the present invention.

Claims

1. A method for evaluating urban road congestion considering macro-micro data coupling, characterized in that: Includes the following steps: S1: Obtain GIS geographic information data within the evaluation road network area, and model the traffic network within the evaluation road network area using complex network theory; S2: Preprocess the vehicle GPS trajectory data within the evaluation road network area to extract the speed information of each road segment. The speed information of each road segment includes the current speed v of each road segment within the time interval t. ij (t) and the maximum speed limit Obtain the traffic status index U(T) of the road network area within the time interval T; S2-1. Obtain vehicle GPS trajectory data within the evaluation road network area, perform preprocessing, and obtain speed information for each road segment, i.e., the current speed v of each road segment. ij (t) and the maximum speed limit within the time interval t The maximum speed limit within the time interval t It is 95% of the current speed within the time interval t; S2-2. Based on the speed information of each road segment, the traffic state index U(T) within the time interval T is obtained using formula (1). U(T)=g(w ij (t)),t=t m ,T=t M (1) In the formula, w ij (t) represents road segment e ij The operating state within time interval t is expressed by the formula: Time interval t M Greater than the time interval t m ; S3: Based on the current speed v of each road segment within time interval t. ij (t) and the maximum speed limit The infiltration threshold q within the evaluation road network area is obtained. c (T); S3-1. Based on the current speed v of the road segment within time interval t. ij (t), time interval t M and time interval t m The e of each road segment within time period T can be obtained from formula (2). ij average velocity v ij (T): S3-2. Based on the time period T, each road segment e ij average velocity v ij (T) and maximum speed limit The functional state f of each road segment is obtained from formula (3). ij (T): Where q(T) is the threshold of the road segment during the time interval T, where 0 indicates that the road segment is completely blocked and all vehicles are stationary, and 1 indicates that the road segment is free to flow and all vehicles are traveling at maximum speed. S3-3. Select f from step S3-2 ij The infiltration threshold q for evaluating the road network area is obtained when the second largest connected cluster in the connected cluster formed by road segments with (T) = 1 reaches its maximum value. c (T); S4: Based on the seepage threshold q c (T), to obtain the operating status threshold θ(T) for each road segment; S4-1. Based on the seepage threshold q c (T), the seepage threshold q for evaluating the road network area is obtained from formula (4). c Change in (T) in, To pass through closed road section e ij And measure the updated seepage threshold of the modified evaluation road network area; S4-2. Based on the seepage threshold q c (T) and seepage threshold q c Change in (T) The operational status threshold θ(T) of the evaluated road network area is obtained; θ(T) is determined by the e values ​​of each road segment in the evaluation road network area. ij The dynamic threshold θ of the running state ij (T) composition, θ ij (T) is obtained from formula (5). Where λ is the correction coefficient; S5: Based on the traffic status index U(T) of the evaluated road network area within the time interval T, the operational status index P(t) of the evaluated road network area is obtained. Combined with the operational status threshold θ(T), the system congestion index SCI is obtained. S5-1. Based on the traffic state index U(T) of the road network within the time interval T obtained in step S2-2, obtain the operational state index P(T) of the road network area, where P(t) = U(T); S5-2. Based on the operational status index P(t) of each road segment within the evaluation road network area from 0:00 to 6:00 AM, the free flow range of the evaluation road network area is obtained. The minimum value of the free flow range is the lower bound index P(t) of the free flow range. pre ); S5-3. Based on the operating state index P(t) and the operating state threshold θ(T), the resistance coefficient R is obtained from formula (6). e (t), S5-4. Based on the operating status index P(t) and the lower bound index P(t) of the free flow range. pre The congestion level C can be obtained from formula (7). m (t), S5-5. Based on the drag coefficient R e (t) and congestion level C m (t), from the formula SCI(t) = C m (t)+R e (t) yields the system congestion index SCI.

2. The urban road congestion evaluation method considering macro-micro data coupling according to claim 1, characterized in that: The specific steps of step S1 are as follows: S1-1. Model the traffic network as a graph G(R, E, W), where R = {r1, ..., r2} n } is a set of nodes, where n is the number of nodes, and E = {e} ij Let {i,j=1,...,k} be the set of road segments, and e ij To connect node r i and node r j The road segments, k is the number of road segments, W = {w ij The time weight matrix is ​​defined as {(t), i,j=1,…,k}, where w ij (t) represents road segment e ij The running status within time interval t.

3. The urban road congestion evaluation method considering macro-micro data coupling according to claim 2, characterized in that: Time interval t M The interval is 60 minutes, and the time interval is t. m It lasts for 5 minutes.

4. A method for evaluating urban road congestion considering macro-micro data coupling according to any one of claims 1 to 3, characterized in that: The preprocessing method for vehicle GPS trajectory data in step S2-1 is as follows: the vehicle GPS trajectory data is matched with the road network of the evaluation road network area using a map matching algorithm, and the dense GPS data accumulated at the locations of stationary vehicles or other vehicles that have been parked for a long time is removed using the DBSCAN clustering method.

Images