CN1707544A - Estimation Method of Traffic Flow State in Urban Road Network - Google Patents

Abstract

A method for estimating the state of urban road network traffic flow, focusing on vehicle-mounted GPS satellite positioning data, combined with the traffic signal state information provided by the corresponding Sydney Adaptive Traffic Control System (SCATS), taking unit-directed road sections as the object, and analyzing the urban road network The traffic flow state is modeled by the least squares method in the three-dimensional space of distance, time and speed. Through the conversion of the curved surface on the three-dimensional space to the curve on the two-dimensional space, the average speed of each directional road section in the urban road network along the road section is obtained at a fixed time, and the analysis and estimation of the current traffic flow congestion state is completed using the speed as an index. . Each unit directional road section between two signal lights corresponds to a traffic flow state model, and the traffic flow state estimation of the entire road network is realized by connecting the surface models of adjacent co-directional road sections.

Description

Estimation Method of Traffic Flow State in Urban Road Network

technical field

The invention relates to a method for estimating the traffic flow state of an urban road network, which is used for estimating the road congestion state in an advanced urban traffic management system and belongs to the field of intelligent traffic research.

Background technique

With the rapid development of social economy, on the one hand, the demand for traffic has increased greatly, while the growth of roads has gradually reached its limit, which further intensifies the contradiction between traffic demand and supply; on the other hand, the rapid progress of information technology has brought comprehensive solutions to traffic problems. opportunity. It is against this background that the advanced traffic information management system (ATIMS) has received extensive attention before other systems of the intelligent transportation system (ITS), and has developed rapidly in countries all over the world, and has been applied to dynamic route planning, Dynamic navigation, road network coordination traffic signal system, dynamic traffic dispatching and other aspects. Among them, the dynamic estimation and prediction of real-time road network traffic status is a key component of ATIMS.

The real-time estimation and prediction of road network traffic status is related to the traffic information used. Different traffic information determines the different methods and accuracy of estimation and prediction. At present, there have been many related researches in the world, among which, Martin L. Hazclton ("Estimating Vehicle Speed from Count and Occupancy data", Journal of Data Science 2(2004), 231-244) is representative according to the road detection ring data. Vehicle flow and duty cycle information are studied using Markov Chain Monte Cal theory. Martin L. Hazclton effectively considered and modeled the problem of high error rate and low reliability of road detection ring data, and the results are gratifying, but his research was carried out on the highway, which is only applicable to the situation where the traffic flow is a continuous flow , and the urban traffic flow is discontinuous flow, which is not suitable for the estimation of the traffic flow state of the urban road network. Using the detection ring data to estimate the traffic flow of the urban road network has high requirements on the urban infrastructure. In many cities, enough required data is often not obtained, and the problem of high error rate cannot be effectively solved.

Contents of the invention

The purpose of the present invention is to address the deficiencies of the prior art and propose a new traffic flow state estimation method for urban road networks, which has the advantages of simple calculation, good real-time performance, and low dependence on urban infrastructure conditions.

In order to achieve such purpose, the technical solution of the present invention focuses on the vehicle-mounted data provided by the Global Positioning System (GPS) with advantages such as high precision, large amount of data, and wide distribution in the city, combined with the corresponding Sydney self-adaptive traffic control system (SCATS) provides stable traffic signal status information to estimate road congestion status. The SCATS system is an adaptive control system for urban traffic lights, which can provide information including vehicle flow, vehicle duty cycle data, and traffic light configuration data at intersections. Among them, the vehicle flow and vehicle duty cycle data still depend on the detection ring buried underground at the intersection, but the status information of traffic lights does not have this dependence, which has the advantage of being stable and accurate.

In the method of the present invention, the directional road section between two signal lights in the urban road network is regarded as a processing unit, and the vehicle-mounted GPS data points that can provide information including distance, time, speed, etc. are regarded as traffic flow on each directional road section Sampling points for a period of time T in the three-dimensional space with distance, time and speed as the coordinate axis to carry out surface fitting modeling to obtain the speed distribution of traffic flow in the space-time domain, on this basis, Obtain the speed change curve of the traffic flow along the road section at a certain moment, so as to obtain the average speed of the road section at this moment, and use this as an index to estimate the traffic congestion state of the road section. The average speed of road sections in the road network is divided into five speed levels, corresponding to five road congestion states: smooth, relatively smooth, not smooth, congested, and severely congested. The estimation of the traffic flow state of the road network is carried out on the basis of the unit directed road segment. Each unit directional road section between two signal lights corresponds to a traffic flow state model, and the traffic flow state estimation of the entire road network is realized by connecting the surface models of adjacent co-directional road sections.

The inventive method mainly comprises the following steps:

1. Data preprocessing for GPS data:

The on-board GPS data points that can provide information including distance, time, and speed are regarded as sampling points of traffic flow on each directional road segment. The preprocessing of GPS data is mainly aimed at the vehicle sampling points with a speed of 0 on the road section. These points consist of two parts: the points where the speed is 0 due to severe congestion and the points where the speed is 0 due to a red signal light. The former is the key part in the analysis of road congestion status, and the latter is the interference part. Judge each sampling point according to the status information of the signal light provided by the SCATS system to see if the sampling point is a point with a speed of 0, and if so, check whether the corresponding time belongs to the red light period of the corresponding signal light represented by the SCATS system, if are satisfied, the sampling point is removed from the set of data points to be fitted.

2. Modeling traffic flow of unit-directed road sections:

The urban road network is formed by connecting intersections and road sections. Traffic control lights are set at major intersections. These signal lights isolate the road sections. The unit directed road section is regarded as a processing unit, which is the basic research object model.

Taking the GPS data sampling points on the unit directional road section within the time period T as the object, and taking the polynomial function space as the basic model space, first select the corresponding polynomial model according to the number of valid data points to be fitted (in the form of The number of several points to be fitted is the threshold value, when it is greater than this threshold value, select the bicubic polynomial for processing, and when it is less than this threshold value, use its degenerate form), and then these sampling points in the three-dimensional space of distance, time and speed The least square method is used to fit the modeling, and the speed distribution surface model of the unit directed road section in time and space is obtained in the T time period.

3. Urban road network traffic flow modeling:

The two road sections are connected to each other at the intersection. Due to the role of traffic lights, the traffic behavior at the intersection is very complicated, and the traffic on adjacent road sections affects each other. Therefore, starting from the traffic flow of the entire road network, when estimating the state of traffic flow, it is necessary to increase the boundary conditions describing the traffic conditions at intersections to consider the influence of adjacent road sections. Based on the speed distribution surface model of the unit directional road section obtained in the second step, let the distance variable be the maximum length of the road section, and obtain the speed change curve of the road section and its adjacent co-directional road section in the time period T. A limited number of points are selected on this curve, and these points are used as the boundary conditions of the velocity distribution surface model corresponding to the adjacent co-directional road section of this directional road section, and participate in the adjacent co-directional road section together with the GPS data sampling points. Fitting of road section speed distribution surface model.

The velocity distribution surfaces of the directional road sections of two co-directional units are connected to each other, and finally the speed distribution of the entire urban road network in time and space is obtained. Among them, each directional road section in the urban road network corresponds to a velocity distribution surface model in time and space.

4. Calculate the average speed of the road section:

For a unit directed road section, on the basis of the mathematical expression of its speed distribution surface model, let the time variable be a constant value t ₀ in the time period T, and obtain the speed of the unit directed road section along the road direction at time t ₀ distribution curve. Integrate this speed distribution curve in the direction of the road to get the average speed of the unit in the direction of the road section at time t ₀ . Calculate the average speed of each directional road section in the road network one by one, and obtain the average speed of each directional road section in the urban road network at time t ₀ .

5. Predict road congestion status from average speed

The road congestion state is estimated by taking the average speed of each directional road section in the urban road network at time _t0 as an index. The average speed of road sections in the urban road network is divided into five speed grades, which correspond to five road congestion states: smooth, relatively smooth, not smooth, congested, and severely congested. The congestion state of each directional road section is judged according to the speed level of the average speed corresponding to each directional road section.

The invention effectively overcomes the dependence of general traffic flow estimation methods on urban hardware facilities, avoids the problems of insufficient and low reliability of general urban traffic flow monitoring facilities, and has the advantages of simple calculation, fast calculation speed and high reliability.

Description of drawings

Figure 1 is a block diagram of the method for estimating the state of urban road network traffic flow.

Fig. 2 is a schematic diagram of data preprocessing for GPS data.

Figure 3 is a schematic diagram of the transformation of the model form.

Figure 3 takes two adjacent road sections in the same direction as an example, showing the whole process of the data form in the process of urban road network traffic flow state estimation from point to plane and plane to line. Among them, Figure 3(a) is a schematic diagram of GPS data sampling points on a unit directed road section; Figure 3(b) is a schematic diagram of a unit directed road section velocity distribution surface model; Figure 3(c) is a velocity distribution surface model of two adjacent road sections Schematic diagram of the connection; Figure 3(d) Schematic diagram of the speed change curve along the direction of the road section at a fixed time.

Figure 4 is a schematic diagram of traffic flow state estimation in the traffic road network of Xuhui District, Shanghai.

Detailed ways

In order to better explain the technical solution of the present invention, a further detailed description will be made below in conjunction with the accompanying drawings and embodiments.

The input data required by the present invention is vehicle satellite positioning data of the GPS system, which provides dynamic traffic detection information including sampling vehicle number, time, position, speed, running direction, and vehicle status. Auxiliary information is the status information of the signal lamp provided by the SCATS traffic adaptive control system, including the phase of the signal lamp and the phase conversion cycle.

The present invention adopts the urban road network traffic flow state estimation scheme shown in Fig. 1, and concrete implementation steps are as follows:

1. Data preprocessing for GPS data

The on-board GPS data points that can provide information including distance, time, and speed are regarded as sampling points of traffic flow on each directional road segment. The preprocessing of GPS data is mainly aimed at the vehicle sampling points with a speed of 0 on the road section. These points consist of two parts: the points where the speed is 0 due to severe congestion and the points where the speed is 0 due to a red signal light. The former is the key part in the analysis of road congestion status, and the latter is the interference part. In general, the distribution of vehicles on the unit's directed road section is shown in Figure 2. l is the trunk of the road, and L is the total length of the road. Along the direction of the road section, the signal light B at the front end of the road section is the main research object. Since the signal light B is in a red light period, a caravan waiting for the red light is formed on the road section, and the vehicle speed v=0. These vehicle points are redundant points to be removed in data preprocessing.

Judge each sampling point according to the status information of the signal light provided by the SCATS system to see if the sampling point is a point with a speed of 0, and if so, check whether the corresponding time belongs to the red light period of the corresponding signal light represented by the SCATS system, if are satisfied, the sampling point is removed from the set of data points to be fitted. That is: for the vehicle sampling point P _i (l _i , t _i , v _i ) on the unit directional road section, if v _i =0, and t _i ∈ T _red (T _red is the red light cycle time of the corresponding signal light), The sampling point is considered as a redundant point, and it is removed from the data point set to be fitted.

2. Modeling traffic flow of unit-directed road sections

The urban road network is formed by connecting intersections and road sections. Traffic control lights are set at major intersections. These signal lights isolate the road sections. The unit directed road section is regarded as a processing unit, which is the basic research object model.

Taking the GPS data sampling points on the unit directional road section in the time period T as the object, the least square method is used to fit the model in the three-dimensional space of distance, time and speed. Figure 3(a) shows the GPS data sampling points distributed on the directed road section 2, which constitute a set of sampling points

When modeling, the polynomial function space is used as the basic model space. The degree of the corresponding polynomial model is selected according to the number of valid data points to be fitted. The number of 20 points to be fitted is taken as the threshold, and when it is greater than this threshold, bi-cubic polynomial is used for processing, and when it is less than this threshold, its degenerate form is used. Right now: $v_Q=f_Q(l,t)=\underset{m=0}{\overset{a}{\mathrm{\Σ}}}\underset{\mathrm{no}=0}{\overset{b}{\mathrm{\Σ}}}{a}_{\mathrm{mn}}{l}^m{t}^{\mathrm{no}},$ Among them: a _mn is the parameter to be sought; l is the distance between the sampling point and the starting point of the road; t is the moment at which the sampling point is located; v _LQ is the speed distribution on the road section within the time period T; (a, b) is to be fitted The highest degree of the polynomial model; Mdata is the number of GPS data sampling points on the directed road section, satisfying:

$<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; a</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; b</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\left\{\begin{array}{cc}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 3,3</span>}<span\; class="notranslate">\; )</span>& \mathrm{<span\; class="notranslate">\; Mdata</span>}<span\; class="notranslate">\; \&Greater\; Equal;</span>\mathrm{<span\; class="notranslate">\; 20</span>}\\ <span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2,2</span>}<span\; class="notranslate">\; )</span>& \mathrm{<span\; class="notranslate">\; 10</span>}<span\; class="notranslate">\; \&le;</span>\mathrm{<span\; class="notranslate">\; Mdata</span>}<span\; class="notranslate">\; <</span>\mathrm{<span\; class="notranslate">\; 20</span>}\\ <span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1,1</span>}<span\; class="notranslate">\; )</span>& \mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; <</span>\mathrm{<span\; class="notranslate">\; Mdata</span>}<span\; class="notranslate">\; \&le;</span>\mathrm{<span\; class="notranslate">\; 10</span>}\\ <span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1,0</span>}<span\; class="notranslate">\; )</span>& \mathrm{<span\; class="notranslate">\; Mdata</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 2</span>}\\ <span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 0,0</span>}<span\; class="notranslate">\; )</span>& \mathrm{<span\; class="notranslate">\; Mdata</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}\end{array}\right.$

Obtain the speed distribution surface model of the unit directional road segment in time and space in T time period. Figure 3(b) shows the fitting results.

3. Urban road network traffic flow modeling

The surface models of the co-directional road sections of each unit are connected to each other to form a road network. Based on the speed distribution surface model of the unit directional road section obtained in the second step, let the distance variable be the maximum length of the road section, and obtain the speed change curve of the road section and its adjacent co-directional road section in the time period T. A limited number of points are selected on this curve, and these points are used as the boundary conditions of the velocity distribution surface model corresponding to the adjacent co-directional road section of this directional road section, and participate in the adjacent co-directional road section together with the GPS data sampling points. Fitting of road section speed distribution surface model. Figure 3(c) shows the connection of two adjacent road section speed distribution surface models, where surface 2 is the fitting result of directional road section 2 after adding the boundary conditions provided by surface 1.

The traffic flow surface model corresponding to directed road segment 1 is f _LQ1 (l, t), _L1 is the length of directed road segment 1, and Mdata2 is the number of GPS data sampling points on directed road segment 2:

${\mathrm{<span\; class="notranslate">\; v</span>}}_{\mathrm{<span\; class="notranslate">\; Q</span>}\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; Q</span>}\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; )</span>{<span\; class="notranslate">\; |</span>}_{\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; L</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; t</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; ,</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; t</span>}}_{<span\; class="notranslate">\; [</span>\frac{\mathrm{<span\; class="notranslate">\; Mdata</span>}\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; ]</span>}}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; v</span>}}_{\mathrm{<span\; class="notranslate">\; Q</span>}\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; L</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; t</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ,</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; v</span>}}_{\mathrm{<span\; class="notranslate">\; Q</span>}\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; L</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; t</span>}}_{<span\; class="notranslate">\; [</span>\frac{\mathrm{<span\; class="notranslate">\; Mdata</span>}\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; ]</span>}<span\; class="notranslate">\; )</span>$

Transform these data points into:

$<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 0</span>}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; t</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; v</span>}}_{\mathrm{<span\; class="notranslate">\; Q</span>}\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; L</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; t</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; )</span><span\; class="notranslate">\; ,</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; ,</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 0</span>}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; t</span>}}_{<span\; class="notranslate">\; [</span>\frac{\mathrm{<span\; class="notranslate">\; Mdata</span>}\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; ]</span>}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; v</span>}}_{\mathrm{<span\; class="notranslate">\; Q</span>}\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; L</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; t</span>}}_{<span\; class="notranslate">\; [</span>\frac{\mathrm{<span\; class="notranslate">\; Mdata</span>}\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; ]</span>}<span\; class="notranslate">\; )</span>$

Taking these data points as boundary conditions, participate in the fitting of the traffic flow surface model of the directed road section 2.

The velocity distribution surfaces of the directional road sections of two co-directional units are connected to each other, and finally the speed distribution of the entire urban road network in time and space is obtained. Among them, each directional road section in the urban road network corresponds to a velocity distribution surface model in time and space.

4. Calculate the average speed of the road section

For a unit directed road section, on the basis of the mathematical expression of its speed distribution surface model, let the time variable be a constant value t ₀ in the time period T, and obtain the speed of the unit directed road section along the road direction at time t ₀ distribution curve. Integrate this speed distribution curve in the direction of the road to get the average speed of the unit in the direction of the road section at time t ₀ .

f _LQ2 (l, t) is the traffic flow surface model corresponding to directed road section 2. f _LQ2 (l, t ₀ ) is the speed change curve of traffic flow along the road section at time t ₀ . When t ₀ =3, the change curve of traffic flow speed along the direction of road section 2 on directional road section 2 is shown in Fig. 3(d).

The average speed is: ${\tilde{v}}_2=\frac{{\&Integral;}_0^{L_2}\underset{m=0}{\overset{a}{\mathrm{\&Sigma;}}}\underset{\mathrm{no}=0}{\overset{b}{\mathrm{\&Sigma;}}}{a}_{\mathrm{mn}}{l}^m{t_0}^{\mathrm{no}}\&Center\; Dot;\mathrm{dl}}{L_2}$

Calculate the average speed of each directional road section in the road network one by one, and obtain the average speed of each directional road section in the urban road network at time t ₀ .

5. Predict road congestion status from average speed

The road congestion state is estimated by taking the average speed of each directional road section in the urban road network at time _t0 as an index. The average speed of road sections in the road network is divided into five speed grades, which correspond to five road congestion states: smooth, relatively smooth, unsmooth, congested, and severely congested, which can be represented by different colors (Figure 4). The congestion state of each road section is judged according to the speed level of the average speed corresponding to each directional road section.

Claims (1)

1, a kind of city road network traffic flow state estimation method, it is characterized in that comprising the steps: 1) The global satellite positioning system data points that can provide information including distance, time, and speed are regarded as sampling points of traffic flow on each directional road section, and each sampling point is judged according to the signal light status information provided by Sydney adaptive traffic control system, Check whether the sampling point is a point with a speed of 0, and if so, check whether its corresponding moment belongs to the red light period of the corresponding signal light represented by the Sydney adaptive traffic control system. If all are satisfied, remove the sampling point from the The data points are removed centrally; 2) Taking the global satellite positioning system data sampling points on the unit directional road section in the time period T as the object, using the bicubic polynomial and its degenerated form as the basic model modeling, firstly according to the number of effective data points to be fitted to select the degree of the corresponding polynomial model, and then use the least squares method to fit and model these sampling points in the three-dimensional space of distance, time and speed, and obtain the speed distribution surface model of the unit's directed road section in time and space in the T time period ; 3) On the basis of the speed distribution surface model of the unit directional road section, let the distance variable be the maximum length of the road section, and obtain the speed change curve of the joint between the road section and its adjacent co-directional road section in the time period T, and take a fixed value on this curve A limited number of points, these points are used as the boundary conditions of the velocity distribution surface model corresponding to the adjacent co-directed road section of this directional road section, and participate in the velocity distribution of the adjacent co-directional road section together with the GPS data sampling points The fitting of the surface model connects the velocity distribution surfaces of the directional road sections of two co-directional units to each other, and finally obtains the speed distribution of the entire urban road network in space and time, wherein each directional road section in the urban road network corresponds to a space-time Velocity distribution surface model; 4) For a unit directed road section, on the basis of the mathematical expression of its speed distribution surface model, let the time variable be a constant value t ₀ in the time period T, and obtain the time t ₀ of the unit directed road section along the road direction The speed distribution curve of the speed distribution curve is integrated in the direction of the road to obtain the average speed of the unit in the direction of the directional road section at time t ₀ , and the average speed of the road section is calculated for each directional road section in the road network one by one to obtain t ₀ The average speed of each directional road section in the urban road network at any time; 5) Taking the average speed of each directional road section in the urban road network at time _t0 as an index to estimate the road congestion state, the average speed of the road section in the urban road network is divided into five speed grades, corresponding to smooth, relatively smooth, There are five road congestion states: unblocked, congested, and severely congested. According to the speed grade of the average speed corresponding to each directional road segment, the congestion state of each directional road segment is judged, and finally the state estimation of the traffic flow of the entire urban road network is obtained.

Images