A unified representation method of highway traffic flow indicators based on space-time correlation

Abstract

This invention discloses a unified expression method for highway traffic flow indicators based on spatiotemporal correlation. The specific steps include: acquiring highway traffic flow indicator data using IoT monitoring equipment; dividing the highway network and spatial information into traffic flow segments and determining the time scale; simplifying the traffic flow indicator data into a speed-vehicle count model and storing it in different time and spatial dimensions to establish a basic data model; calculating flow rate and density under different spatiotemporal conditions based on the speed-vehicle count model and performing statistical analysis to construct a practical flow density model and an optimized theoretical model; storing the flow density theoretical model in different spatial dimensions and cross-validating it with the practical model. This invention retains all basic data, reflects the spatiotemporal correlation between data and the relationships between data indicators, efficiently expresses traffic flow data and indicators, reduces data lookup time, and establishes an effective data layer model for data prediction.

Description

Technical Field

[0001] This invention belongs to the field of traffic monitoring technology, specifically relating to a unified expression method for highway traffic flow indicators based on spatiotemporal correlation. Background Technology

[0002] Highways generate massive amounts of traffic data from multiple sources and heterogeneous structures. For example, dynamic data includes data collected from tollbooth card swipes, ETC gantries, surveillance cameras, and GPS data of key vehicles—data collected by various traffic sensors. Static data includes highway network data, entrances and exits, interchanges, road sections, tunnels, and service areas. Traffic indicator data includes flow rate, speed, density, congestion mileage, travel time, and the evolution of traffic indicators over different time periods. This data is characterized by its diverse types, varied structures, and complex spatiotemporal relationships. However, existing highway data representation and modeling methods are inefficient, complex, and lack strong correlations between data points, resulting in wasted computational resources and hindering effective real-time assessment of traffic conditions, thus impacting urban development.

[0003] Therefore, existing technologies suffer from the technical problem of low correlation between multi-source heterogeneous traffic data and long search times. Summary of the Invention

[0004] To overcome the shortcomings of existing technologies, this invention provides a unified expression method for highway traffic flow indicators based on spatiotemporal correlation, so as to reduce traffic data search time and establish an effective data layer model for data prediction.

[0005] The technical problem solved by this invention can be achieved by the following technical solutions:

[0006] The aforementioned unified expression method for highway traffic flow indicators based on spatiotemporal correlation includes the following specific steps:

[0007] S1. Obtain highway traffic flow index data based on IoT monitoring equipment;

[0008] S2. Based on the highway network and spatial information, delineate the pile ends and determine the time scale;

[0009] S3. Simplify traffic flow index data into a speed-vehicle count model and store it in different time and space dimensions to establish a basic data model;

[0010] S4. Calculate the flow rate and density under different time and space conditions based on the speed-vehicle number model, and perform statistical analysis to construct a practical model and an optimized theoretical model of flow rate and density.

[0011] S5. Store the flow density theoretical model in different spatial dimensions and verify it with the actual model.

[0012] Furthermore, in step S2, the specific operation is as follows:

[0013] S21. Divide each road into multiple equal-length pile segments, with a pile segment length of L. Set a large time period T and a small time period t according to the pile segment length, and adjust the size of the time period as needed.

[0014] S22. Establish a spatial coordinate system T1, i.e., the spatial relationship layer, using road names and pile segments as the horizontal and vertical coordinate axes;

[0015] S23. Based on the pile segment length L, set a large time period T and a small time period t. In order to effectively analyze the time correlation of the data, use the small time period t and the large time period T as the horizontal and vertical axes respectively to establish a time coordinate system T2, that is, the time relationship layer.

[0016] Furthermore, in step S3, the specific steps are as follows:

[0017] S31. Simplify the high-dimensional vector constructed from traffic flow index data into speed and its corresponding number of vehicles;

[0018] S32. Based on the statistically obtained vehicle speed data, establish a speed-vehicle count model for each pile segment at different times;

[0019] S33. Store the time coordinate system T2 at each discrete point in the spatial coordinate system T1 to establish a basic data model to represent the traffic flow data of a certain road segment at various times.

[0020] Furthermore, the step of establishing the speed-vehicle count model in step S32 is as follows: calculate the average speed of each vehicle, and establish a basic data coordinate system T3 with the average speed and vehicle count as the horizontal and vertical axes. Store the basic data coordinate system T3 at each discrete point of the time coordinate system T2 to represent the speed and corresponding vehicle count data of a certain road segment at a certain time.

[0021] Further, in step S4, based on the basic data coordinate system T3, the flow rate q at this pile segment is calculated from the total number of vehicles, the density ρ is calculated from the number of vehicles and the length L of the pile segment, and the average speed v is calculated from the speed v and the corresponding number of vehicles. A flow rate density relationship coordinate system T4 is established with density ρ and flow rate q as the horizontal and vertical axes, and the flow rate and density at each time in the pile segment are represented in the flow rate density relationship coordinate system T4 in the form of scattered points to represent the flow rate density correspondence distribution map of a certain pile segment of a certain road.

[0022] Furthermore, step S4 specifically includes the following operations:

[0023] S41. Store the speed-vehicle count model as a node in its corresponding spatiotemporal relation layer.

[0024] S42. Based on the speed-vehicle count model, the average speed at this pile segment is calculated:

[0025]

[0026] In the formula, the average speed is in km / h, and the speed of each vehicle is taken as the midpoint of the speed range, excluding low-speed and speeding vehicles.

[0027] S43. Based on the speed-vehicle count model, the flow rate at this pile segment is calculated: The unit is vehicles / h, and the road section used for the flow rate is the section at the center of the pile segment;

[0028] S44. Based on the speed-vehicle number model, the density at this pile segment is calculated as: ρ=q / L, in units of vehicles / km;

[0029] S45. Based on the calculated velocity-density flow index, plot a scatter plot. Using the Underwood flow-density model in the flow-density relationship coordinate system T4, plot the fitted curve of the flow-density relationship. The formula is:

[0030]

[0031] In the formula, k is the density, and u f To ensure smooth traffic flow, k m For optimal density;

[0032] S46. Optimize the parameters of the relational analysis model using actual traffic flow indicators. Improve the fit between the Underwood flow density model and the actual traffic flow indicators by adjusting the parameters in the Underwood flow density model.

[0033] Furthermore, the specific operation of step S5 is as follows: the flow density relationship coordinate system T4 with the Underwood fitting curve is stored in the spatial coordinate system T1. The flow density relationship coordinate system T4 can be mutually verified with the basic data coordinate system T3. When there is a large deviation between the data in the flow density relationship coordinate system T4 and the actual data, specifically when the flow density data of multiple time dimensions of the same road segment show a large deviation, the parameters of the Underwood flow density model should be modified according to the actual data to improve the model's fitting degree. In addition to the case of large deviation from the actual data, the Underwood flow density model should also be updated regularly.

[0034] Compared with existing technologies, the present invention has the following advantages: The present invention processes existing sensor data and static data, and performs dimensionality reduction processing on multi-dimensional index data, simplifying multiple traffic flow indicators into a relationship model between speed and the corresponding number of vehicles. The time period can be adjusted in the time relationship layer, and the characteristics of traffic flow data under different time periods are analyzed. This expression method retains all basic data and reflects the spatiotemporal correlation between data and the relationship between data indicators. It efficiently expresses highway traffic flow data and indicators, reduces data search time, and establishes an effective data layer model for data prediction. Attached Figure Description

[0035] Figure 1 This is a flowchart of the method of the present invention;

[0036] Figure 2 This is a schematic diagram of the speed-vehicle number model of the present invention;

[0037] Figure 3 This is a schematic diagram of the spatial model of the present invention;

[0038] Figure 4 This is a schematic diagram of the time model of a certain pile section of a highway in an embodiment of the present invention;

[0039] Figure 5 This is a schematic diagram of the speed-vehicle count model for the first Monday of the first week on a highway, as described in an embodiment of the present invention.

[0040] Figure 6 This is a scatter plot of the traffic flow and density indicators of this invention.

[0041] Figure 7 This is a schematic diagram of the fitting curve of the Underwood flow density model in an embodiment of the present invention. Detailed Implementation

[0042] The following description, in conjunction with the accompanying drawings, further illustrates the unified expression method for highway traffic flow indicators based on spatiotemporal correlation according to the present invention.

[0043] This embodiment takes the area from K0+000 to K3+000 of the Liangguang Expressway as an example. This area contains multi-source heterogeneous data such as toll station data, gantry ETC data, surveillance camera data, and GPS data. The following methods are used to express the traffic flow data and indicators in this area and to build a model.

[0044] like Figure 1 As shown, a unified expression method for highway traffic flow indicators based on spatiotemporal correlation includes the following steps:

[0045] (i) Obtain highway traffic flow index data based on IoT monitoring equipment.

[0046] It connects various sensor data such as toll stations, ETC gantries, surveillance cameras, and GPS data of key vehicles on the road, as well as static data such as highway network, entrances and exits, interchanges, pile sections, tunnels, and service areas, and converts the sensor data into usable traffic flow indicator data.

[0047] (ii) Based on the highway network and spatial information, the pile ends are divided and the time scale is determined.

[0048] Based on the locations of sensors such as toll stations, ETC gantries, and surveillance cameras, the road is divided into multiple equal-length pile segments, each with a length of L. The length L of each pile segment should not be too large. A spatial coordinate system T1, i.e., the spatial relationship layer, is established using the road name and pile segment as the horizontal and vertical axes, respectively.

[0049] Based on the length L of the pile segment, a large time period T and a small time period t are set (the size of the time period can be adjusted as needed). To effectively analyze the temporal correlation of the data, the large time period T and the small time period t should not be too disparate. A time coordinate system T2, i.e., the time relationship layer, is established using the small time period t and the large time period T as the horizontal and vertical axes, respectively. The large and small time periods in coordinate system T2 can be parameterized according to the temporal correlation of the road data to achieve the analysis of road traffic flow characteristics under multiple time periods. When the large and small time periods are modified, the data changes in the time coordinate system T2 under each spatial dimension are achieved by changing the address. In this embodiment, the length of the pile segment from K0+000 to K3+000 on the Liangguang Expressway is set to 500m, and the large and small time periods T are weeks and t is days.

[0050] In this application, the data model uses large and small time periods to perform similarity analysis on data between months within a year, between weeks within a month, and between days within a week. The time period is adjusted based on the need for time correlation, unifying data from different time scales into a time relationship model for correlation analysis. The size of the time period can be adjusted according to the data correlation analysis requirements. Changing the time period size requires a corresponding change in the data storage location. Specifically, the storage address of the speed-vehicle-count model is used to represent the speed-vehicle-count model, and this storage address is stored in the time model. Modifying the time period size in the time model only requires changing the storage address at different times.

[0051] (iii) Simplify traffic flow index data into a speed-vehicle count model and store it in different time and space dimensions to establish a basic data model.

[0052] The high-dimensional vector constructed from traffic flow index data is simplified into speed and its corresponding vehicle count data. The average speed of each vehicle is calculated, and a basic data coordinate system T3 is established with the average speed and vehicle count as the horizontal and vertical axes. The basic data coordinate system T3 is stored at discrete points in the time coordinate system T2 to represent the speed and corresponding vehicle count data of a certain road segment at a certain time. The time coordinate system T2 is stored at discrete points in the spatial coordinate system T1 to establish a basic data model to represent the traffic flow data of a certain road segment at various times.

[0053] In this embodiment, data from various IoT monitoring devices, such as toll stations, gantry ETC, surveillance cameras, and GPS devices for key vehicles, are calculated, converted into vehicle speed data, and statistically analyzed, as shown in the table below:

[0054]

[0055] For example, the vehicle speed data for the first day of the first week at the K0+500 to K1+000 section of the Liangguang Expressway is statistically analyzed as {A1, A2, A3, A4, A5, A6, A7, A8} = {1, 72, 144, 286, 197, 136, 39, 3}. Based on this statistical data, speed-vehicle count models are established for each section at different times. The speed-vehicle count model for the K0+500 to K1+000 section on the first day of the first week is as follows: Figure 2 As shown.

[0056] like Figure 3 and Figure 4 As shown, a spatial model and a temporal model are established based on the pile segment length L, the large time period T, and the small time period t, and are respectively divided into a spatial relationship layer and a temporal relationship layer, wherein the temporal model is stored as a node in the spatial model. Figure 3 At point p1 in the middle, the horizontal road is the Liangguang Expressway, and the vertical segment is from K0+500 to K1+000. p1 represents the storage address of the time relationship layer model for the Liangguang Expressway from K0+500 to K1+000. Similarly, Figure 4 The address q2 represents the storage address of the speed-vehicle count model for the section from K0+500 to K1+000 of the Liangguang Expressway on Monday of the first week.

[0057] (iv) Calculate the flow rate and density in different time and space according to the speed-vehicle number model, and perform statistics to construct the actual flow rate and density model and the optimization theoretical model.

[0058] Based on the basic data coordinate system T3, the flow rate q at this pile segment is calculated from the total number of vehicles, the density ρ is calculated from the number of vehicles and the length L of the pile segment, and the average speed v is calculated from the speed v and the corresponding number of vehicles. A flow rate density relationship coordinate system T4 is established with density ρ and flow rate q as the horizontal and vertical axes, and the flow rate and density at each time in this pile segment are represented as scattered points in the flow rate density relationship coordinate system T4 to represent the flow rate density distribution map of a certain pile segment of a certain road.

[0059] like Figure 5 As shown, the speed-vehicle count model is stored as a node in its corresponding spatiotemporal relation layer. Based on the speed-vehicle count model, the average speed at this pile segment is calculated:

[0060]

[0061] In the formula, the average speed is in km / h, and the speed of each vehicle is taken as the midpoint of the speed range. Low-speed and speeding vehicles are not calculated.

[0062] Based on the speed-vehicle-number model, the flow rate at this pile segment was calculated: The unit is vehicles / h, and the road section used for the flow rate is the section at the center of the pile segment.

[0063] Based on the speed-vehicle number model, the density at this pile segment is calculated as: ρ = q / L, with units of vehicles / km.

[0064] A scatter plot was drawn based on the calculated velocity-density traffic flow index. The Underwood flow-density model was then used to plot the fitted curve of the flow-density relationship in the flow-density coordinate system T4. The formula is as follows:

[0065]

[0066] In the formula, k is the density, and u f To ensure smooth traffic flow, k m The optimal density is determined as follows: In this embodiment, the average vehicle speed on Monday of the first week at the section from K0+500 to K1+000 on the Liangguang Expressway is approximately 88.4 km / h; the traffic flow is 878 vehicles / h; and the density is 1756 vehicles / km.

[0067] The parameters of the relationship analysis model were optimized using actual traffic flow indicators. By adjusting the parameters in the Underwood flow density model, the fit between the Underwood flow density model and actual traffic flow indicators was improved. Figure 7 As shown, the optimized Underwood flow density model has a free-flowing speed of 80 km / h and an optimal density of 227 vehicles / km.

[0068] (v) Store the flow density theoretical model in different spatial dimensions and verify it with the actual model.

[0069] The flow density relationship coordinate system T4 with the Underwood fitted curve is stored in the spatial coordinate system T1. The flow density relationship coordinate system T4 can be mutually verified with the basic data coordinate system T3. When there is a large deviation between the data in the flow density relationship coordinate system T4 and the actual data, specifically when the flow density data of multiple time dimensions of the same road segment show a large deviation, the parameters of the Underwood flow density model should be modified according to the actual data to improve the model's fit. In addition to the case of large deviation from the actual data, the Underwood flow density model should also be updated regularly.

[0070] In one specific embodiment, the density of a certain section of the Liangguang Expressway at a certain time is 400 vehicles / km. At this time, the theoretical flow rate in the Underwood flow density model should be 4000 vehicles / h. If the actual flow rate is 3892 vehicles / h, it is considered that the Underwood flow density model has a high degree of fit with the actual traffic flow index. If the actual flow rate is 3203 vehicles / h, it is necessary to compare the actual flow rates at multiple time points to check the randomness of the data. If it is not random data, the model parameters need to be adjusted to improve the fit between the model and the actual traffic flow index.

[0071] The permissible error between theoretical and actual data should be defined based on the characteristics of historical road data.

[0072] This invention utilizes IoT monitoring equipment to acquire highway traffic flow data. Based on spatial information such as the highway network and entrances / exits, it divides the data into segments and determines the time scale. Traffic indicators such as flow velocity and density are simplified into speed-vehicle count models, which are stored in different temporal and spatial dimensions to establish a basic data model. The speed-vehicle count model calculates flow rate and density under different temporal and spatial conditions and performs statistical analysis. This constructs a practical flow density model and an Underwood flow density relationship theory model. Both the practical and theoretical models are stored in different spatial dimensions and cross-validated with actual data. The simplification of traffic flow indicators such as flow rate, speed, and density into speed-vehicle count models, and the storage of these models in a temporal relationship layer, improves data representation efficiency, reflects the spatiotemporal correlation between data and the correlation between indicators, and provides an efficient representation of traffic flow data and indicators.

[0073] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A unified expression method for highway traffic flow indicators based on spatiotemporal correlation, characterized in that, The specific steps include: S1. Obtain highway traffic flow index data based on IoT monitoring equipment; S2. Based on the highway network and spatial information, delineate the pile ends and determine the time scale; S21. Divide each road into multiple equal-length pile segments, with a pile segment length of L. Set a large time period T and a small time period t according to the pile segment length, and adjust the size of the time period as needed. S22. Establish a spatial coordinate system T1, i.e., the spatial relationship layer, using road names and pile segments as the horizontal and vertical coordinate axes; S23. Based on the pile segment length L, set a large time period T and a small time period t. In order to effectively analyze the time correlation of the data, and use the small time period t and the large time period T as the horizontal and vertical axes respectively, establish a time coordinate system T2, that is, the time relationship layer. S3. Simplify traffic flow index data into a speed-vehicle count model and store it in different time and space dimensions to establish a basic data model; S31. Simplify the high-dimensional vector constructed from traffic flow index data into speed and its corresponding vehicle count data; S32. Based on the statistical vehicle speed data, establish a speed-vehicle count model for each pile segment at different times; the steps to establish the speed-vehicle count model are as follows: statistically analyze the average speed of each vehicle, and establish a basic data coordinate system T3 with the average speed and vehicle count as the horizontal and vertical axes. Store the basic data coordinate system T3 at each discrete point of the time coordinate system T2 to represent the speed and corresponding vehicle count data of a certain pile segment of a certain road at a certain time. S33. Store the time coordinate system T2 at each discrete point in the spatial coordinate system T1 to establish a basic data model to represent the traffic flow data of a certain road segment at various times; S4. Calculate the flow rate and density under different time and space conditions based on the speed-vehicle number model, and perform statistical analysis to construct a practical model and an optimized theoretical model of flow rate and density. S41. Store the speed-vehicle count model as nodes in its corresponding spatiotemporal relation layer. S42. Based on the speed-vehicle count model, the average speed at this pile segment is calculated: ， In the formula, the average speed is in km / h, and the speed of each vehicle is taken as the midpoint of the speed range, excluding low-speed and speeding vehicles. S43. Based on the speed-vehicle count model, the flow rate at this pile segment is calculated: The unit is vehicles / h, and the road section used for the flow rate is the section at the center of the pile segment; S44. Based on the speed-vehicle count model, the density at this pile segment is calculated: The unit is vehicles per km; S45. Based on the calculated velocity-density flow index, plot a scatter plot. Using the Underwood flow-density model in the flow-density relationship coordinate system T4, plot the fitted curve of the flow-density relationship. The formula is: ， In the formula, k is the density. To ensure smooth traffic flow, For optimal density; S46. Optimize the parameters of the relational analysis model using actual traffic flow indicators. Improve the fit between the Underwood flow density model and the actual traffic flow indicators by adjusting the parameters in the Underwood flow density model. S5. Store the flow density theoretical model in different spatial dimensions and verify it with the actual model.

2. The unified expression method for highway traffic flow indicators based on spatiotemporal correlation according to claim 1, characterized in that, In step S4, based on the basic data coordinate system T3, the flow rate q at this pile segment is calculated from the total number of vehicles, the density ρ is calculated from the number of vehicles and the pile segment length L, and the average speed is calculated from the speed v and its corresponding number of vehicles. A flow-density relationship coordinate system T4 is established with density ρ and flow rate q as the horizontal and vertical axes, and the flow rate and density at each time in the pile segment are represented as scattered points in the flow-density relationship coordinate system T4 to represent the flow-density relationship distribution map of a certain pile segment of a certain road.

3. The unified expression method for highway traffic flow indicators based on spatiotemporal correlation according to claim 1, characterized in that, The specific operation of step S5 is as follows: store the flow density relationship coordinate system T4 with the Underwood fitting curve into the spatial coordinate system T1. The flow density relationship coordinate system T4 can be mutually verified with the basic data coordinate system T3. When there is a large deviation between the data in the flow density relationship coordinate system T4 and the actual data, specifically when the flow density data of multiple time dimensions of the same road segment show a large deviation, the parameters of the Underwood flow density model should be modified according to the actual data to improve the model's fitting degree. In addition to the case of large deviation from the actual data, the Underwood flow density model should also be updated regularly.

Images