CN115798198A - Urban road network travel time distribution estimation method based on data fusion - Google Patents
Urban road network travel time distribution estimation method based on data fusion Download PDFInfo
- Publication number
- CN115798198A CN115798198A CN202211368233.8A CN202211368233A CN115798198A CN 115798198 A CN115798198 A CN 115798198A CN 202211368233 A CN202211368233 A CN 202211368233A CN 115798198 A CN115798198 A CN 115798198A
- Authority
- CN
- China
- Prior art keywords
- travel time
- path
- rsk
- road network
- distribution
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000009826 distribution Methods 0.000 title claims abstract description 95
- 238000000034 method Methods 0.000 title claims abstract description 40
- 230000004927 fusion Effects 0.000 title claims abstract description 13
- 238000005315 distribution function Methods 0.000 claims description 9
- 239000000126 substance Substances 0.000 claims description 8
- 238000011144 upstream manufacturing Methods 0.000 claims description 8
- 239000000203 mixture Substances 0.000 claims description 7
- 238000009499 grossing Methods 0.000 claims description 6
- 238000013507 mapping Methods 0.000 claims description 4
- 238000007476 Maximum Likelihood Methods 0.000 claims description 3
- 239000000470 constituent Substances 0.000 claims description 3
- 238000005070 sampling Methods 0.000 claims description 3
- 239000000758 substrate Substances 0.000 claims 1
- 230000006870 function Effects 0.000 description 23
- 238000004088 simulation Methods 0.000 description 5
- 238000011156 evaluation Methods 0.000 description 4
- 238000005457 optimization Methods 0.000 description 4
- NAWXUBYGYWOOIX-SFHVURJKSA-N (2s)-2-[[4-[2-(2,4-diaminoquinazolin-6-yl)ethyl]benzoyl]amino]-4-methylidenepentanedioic acid Chemical compound C1=CC2=NC(N)=NC(N)=C2C=C1CCC1=CC=C(C(=O)N[C@@H](CC(=C)C(O)=O)C(O)=O)C=C1 NAWXUBYGYWOOIX-SFHVURJKSA-N 0.000 description 2
- 238000000342 Monte Carlo simulation Methods 0.000 description 2
- 238000012614 Monte-Carlo sampling Methods 0.000 description 2
- 230000008901 benefit Effects 0.000 description 2
- 238000004422 calculation algorithm Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000010354 integration Effects 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 230000002902 bimodal effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000005065 mining Methods 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 230000035515 penetration Effects 0.000 description 1
- 230000035699 permeability Effects 0.000 description 1
- 230000029305 taxis Effects 0.000 description 1
Images
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T10/00—Road transport of goods or passengers
- Y02T10/10—Internal combustion engine [ICE] based vehicles
- Y02T10/40—Engine management systems
Landscapes
- Traffic Control Systems (AREA)
Abstract
The invention provides an urban road network travel time distribution estimation method based on data fusion, which is characterized in that a path rsk is obtained based on network connection vehicle track data uploaded by a floating vehicle, then travel time distribution probability density of any path is calculated, and AVI data calibration time is acquired through an alarm card port device.
Description
Technical Field
The invention relates to the technical field of traffic condition estimation, in particular to an urban road network travel time distribution estimation method based on data fusion.
Background
The travel time and the distribution thereof are important evaluation indexes of the traffic running state of the urban road network, travelers in the urban road network can make more reliable path selection based on accurate estimation of the travel time, the travel time distribution can intuitively show the land benefit and the running reliability of the urban road network, and the travel time distribution can also be used for determining the bottleneck of the road network, thereby applying active control measures and improving the overall traffic condition. In the existing research, a single data source of a floating car or AVI data is generally adopted to estimate the travel time distribution of a road section, and the existing research is mainly oriented to a continuous flow (such as an express way) scene. The floating car is generally a bus and a taxi which are provided with a vehicle-mounted GPS positioning device and run on an urban main road, and the data uploaded by the floating car is the track data of the networked vehicles. However, the problems of low uploading frequency, low market penetration rate, biased samples and the like exist in the online vehicle track data uploaded by the floating vehicle, and the stability of estimation and the accuracy of the estimated travel time distribution cannot be guaranteed; the AVI data (also called vehicle automatic identification data, hereinafter referred to as AVI data) is collected based on the electric police access device, and the arrangement density of the electric police access device in an urban road network is very limited, so that the universality of the travel time distribution estimation based on the AVI data in the road network cannot be guaranteed, and the estimation can be performed only on the road section with better local electric competition device arrangement conditions.
Disclosure of Invention
In order to solve the problems that the existing travel time estimation method is inaccurate or low in universality due to the fact that the stability of floating car data cannot be guaranteed and the coverage rate of electric police monitoring equipment is limited in the prior art, the invention provides the urban road network travel time distribution estimation method based on data fusion, which is not limited by the layout density factor of detectors in a road network and can flexibly and accurately estimate the travel time distribution of any given path in the road network.
The technical scheme of the invention is as follows: a method for estimating the travel time distribution of an urban road network based on data fusion is characterized by comprising the following steps:
s1: determining a road network to be analyzed;
s2: acquiring a road network graph of the road network to be analyzed, wherein the road network graph comprises the flow directions of all roads;
s3: constructing a path unit travel time distribution function;
the path unit is as follows: defining two continuous flow directions in a road network as a path unit;
based on the travel time of the network connection vehicle track data in the road network to be analyzed, fitting initial path unit travel time distribution by using a bivariate Gaussian mixture model GMM, wherein the PDF form of the initial path unit travel time distribution function is given by the following formula:
wherein, t - And t + Representing the travel time of the upstream and downstream flow directions, respectively; c represents the number of component distributions in GMM, λ c Is the weight of component c; tau. c A bivariate gaussian distribution probability density function representing the component distribution c; mu.s c And
is the mean and variance of the component distribution c; a bivariate Gaussian distribution with τ () representing the populationA probability density function;
s4: fitting the initial path unit travel time distribution function GMM through a maximum likelihood function L based on the network connection vehicle track data of each path unit in the road network;
wherein E is a path unit set of all paths included in the road network to be analyzed; n is e The number of the observed internet connected vehicle samples on the path unit E belongs to E; t is t i- And t i+ Representing the travel time of the ith networked vehicle in the upstream and downstream flow directions;
s5: calculating the average travel time t of each flow direction m in the road network to be analyzed m ：
Wherein the content of the first and second substances,
is a mapping relation between the path unit e and the flow m, if the path unit e includes the flow m
Is 1, otherwise
Is 0;
s6: acquiring a target path rsk;
setting: the number of path units included in the target path rsk is NUM;
s7: from the initial path element travel time distribution GMM tau 0 (t - ,t + ) In the random generation of N samples
Initializing a variable nu to 0;
s8: make it possible to
And distributed from the condition
Sampling in the middle;
s9: let nu = nu +1;
s10: circularly executing the steps S8 to S9 until all NUM path units in the target path rsk are listed;
s11: t obtained by approximating sample distribution by using kernel density estimation KDE method rsk Probability density function tau (t) rsk ,b)：
Wherein, t rsk For the mean travel time variable of the path rsk,. Kappa.. Cndot.is a Gaussian kernel, b is a smoothing parameter called the bandwidth, M denotes the flow direction, M rsk Set flow directions contained for path rsk;
s12: constructing an objective function based on AVI data:
wherein P represents a local path, P is a set of paths, RS is a set of OD pairs of traffic start and stop points, K rs Is the path set between OD to rs; z is the function of the object function,
in the objective function, the first term
Is a typical OLS entry;
second item
Is to minimize the variance of the residual error for different paths of the same flow direction combination between p;
w' 1 ,w' 2 the weights of the first term and the second term, respectively, v is the standard deviation of the vehicle travel time between the recorded pair p of AVI data;
s13: solving a minimum objective function minZ based on constraint conditions to obtain a decision variable
The solution of (2);
the following formula is used to determine the variables
The solution of (c) is converted into an adjustment of the average flow direction travel time;
wherein the content of the first and second substances,
indicating the adjustment ratio allocated to the flow direction m on the path rsk in the range of AVI to p; t is t m Is the travel time in flow direction m;
an average travel time adjustment amount indicating the flow direction m;
indicating the occurrence of whether the flow direction m is passed by rsk and within the range of the AVI sequence p; m is a set of all flow directions included in the road network to be analyzed;
s14: an adjustment is assigned to each component of the GMM containing the target flow direction according to the constituent weight:
wherein c is the component in GMM, mu c Is the average value of the component distribution c, lambda c Is the weight of component c;
s15: recalculating t based on the adjusted GMM rsk Probability density function tau (t) rsk And b), obtaining the average travel time of the target path rsk.
It is further characterized in that:
the smoothing parameter b takes the following values:
wherein the content of the first and second substances,
is the sample variance used to perform KDE;
the constraint conditions are as follows:
wherein, t p Represents the average of the recorded travel times between any pair of AVI detectors;
representing a travel time between a target pair of AVIs along path rsk; RS is set of OD pairs, K rs Is the set of paths between the OD versus the rs,
indicates whether the local path p is passed by the path rsk;
is the incidence indicating whether the flow direction m is passed by rsk and is within the scope of the AVI sequence p;
the residual error term is defined as the residual travel time of the path rsk through the AVI sequence p; f. of rsk Representing path traffic data;
in step S12, the default value of the ratio of the two weights is w' 1 :w' 2 ＝1:1。
The invention provides an urban road network travel time distribution estimation method based on data fusion, which comprises the steps of firstly obtaining a path rsk based on network connection vehicle track data uploaded by a floating vehicle, then calculating travel time distribution probability density of any path, and then calibrating time for AVI data acquired by an alarm card port device, wherein in the method, the travel time information locally observed by AVI and the global distribution information provided by a network connection vehicle are combined, so that the problems that network connection vehicle samples which are difficult to deal with in the existing method are biased, and the coverage rate in a road network of vehicle automatic identification equipment is lower are solved, the urban road network travel time distribution estimation is ensured not to be limited by the distribution density factor of a detector in the road network, and the travel time distribution of any given path in the road network can be flexibly and accurately estimated; in the method, the link of the upstream and downstream continuous flow directions is defined as a path unit, and the path unit is used as a basic unit of path travel time distribution, so that more refined modeling is performed, the travel time correlation relation in the modeling unit is simplified, and the requirement on the sample size of the internet vehicle is reduced; the technical route of the invention starts from the flow direction and the travel time distribution of the path units, and generates the sample set of the path unit combination by the Monte Carlo sampling method, therefore, the travel time distribution of any given path in the road network can be flexibly estimated without being limited by factors such as the distribution density of detectors in the road network, and the like, therefore, the method can be widely applied to urban road networks in the future, and has larger potential and wide application range.
Drawings
Fig. 1 is a schematic diagram of a process for estimating travel time distribution of an urban road network in the present application;
FIG. 2 is an embodiment of a path cell;
FIG. 3 is a network satellite map of example 2;
fig. 4 is a schematic diagram of road network topology and AVI distribution in embodiment 2;
FIG. 5 is a VISUM network of example 2;
fig. 6 shows a VISSIM network in embodiment 2.
Detailed Description
The invention comprises an urban road network travel time distribution estimation method based on data fusion, which is characterized by comprising the following steps:
s1: and determining a road network to be analyzed.
S2: and acquiring a road network graph of the road network to be analyzed, wherein the road network graph comprises the flow directions of all roads.
S3: constructing a path unit travel time distribution function;
the path unit is as follows: two successive flow directions in the road network are defined as a path unit. As shown in fig. 2, the upstream is in a straight flow direction, and the downstream is in a left-turn flow direction, so as to form a path unit; the left-turn flow direction and the downstream straight flow direction constitute a path unit.
Based on the definition of the path unit, the vehicle travel path can be characterized by a flow direction sequence or a path unit sequence. By defining path elements, the travel time correlation between successive flow directions can be explicitly estimated.
Determining a bivariate property in view of path unit travel time, fitting an initial path unit travel time distribution using a bivariate gaussian mixture model GMM, the Probability Density Function (PDF) form of the initial path unit travel time distribution Function being given by the following formula:
wherein, t - And t + Represents the travel time of the upstream and downstream flow, respectively; c represents the number of component distributions in GMM, representing the number of modes in the distribution modeling, λ c Is the weight of component c; tau is c A bivariate gaussian distribution probability density function representing the component distribution c; mu.s c And
is the mean and variance of the component distribution c; τ () represents the overall bivariate gaussian distribution probability density function.
In particular, the data samples are fitted (t) - ,t + ) The group number C can be set according to specific requirements, and is usually set to be between 2 and 5.
S4: fitting an initial path unit travel time distribution function GMM by using a maximum likelihood function L based on the network connection vehicle track data of each path unit in the network and based on an EM algorithm;
e is a path unit set of all paths in the road network to be analyzed; n is e The path unit E belongs to the number of the network connection vehicle samples observed on the E; t is t i- And t i+ Representing travel times of the ith networked vehicle in the upstream and downstream flow directions; GMM parameters (λ, μ, σ) can be achieved by employing a classical expectation-maximization algorithm 2 ) And (6) estimating.
S5: calculating the average travel time t of each flow direction m in the road network to be analyzed m ：
Wherein the content of the first and second substances,
mapping relationship between path unit e and flow direction m, if path unit e includes flow direction m
Is 1, otherwise
Is 0;
in general, given path unit travel time distribution, path travel time can be correspondingly estimated through convolution, but because of the difficulty of convolution and integration in charge of GMM, the invention proposes to adopt a Monte Carlo Simulation (MCS) method to approximate path travel time distribution:
deriving a GMM distribution form;
as shown in equation (4), by integration, it can be inferred that the marginal distribution of the bivariate GMM is a univariate GMM, which has the same component number and weight as the corresponding bivariate GMM.
Given a bivariate Gaussian conditional distribution c (t + |t - ) The gaussian distribution after parameter adjustment is still adopted, and the conditional distribution of the GMM is still the gaussian mixture model.
The conditional distribution given by equation (5) is also a univariate GMM whose composition is a weight-normalized univariate gaussian distribution.
Wherein rho is a correlation coefficient and is a fixed parameter in a binary Gaussian distribution probability density function.
Because of the bivariate Gaussian conditional distribution tau given by equation (6) c (t + |t - ) The gaussian distribution after parameter adjustment is still present, while the conditional distribution of the GMM is still a gaussian mixture model.
Then, based on equations (4) to (6), the present application proposes the following steps to generate a sample of an arbitrary path.
S6: acquiring a target path rsk;
setting: the number of path units included in the target path rsk is NUM;
s7: distributing GMM tau from initial path unit travel time 0 (t - ,t + ) In the random generation of N samples
Initializing a variable nu to 0;
s8: make it
And distributed from the condition
Sampling in the middle;
s9: let nu = nu +1;
s10: circularly executing the steps S8 to S9 until all NUM path units in the target path rsk are listed;
s11: t obtained by approximating sample distribution by using kernel density estimation KDE method rsk Probability density function tau (t) rsk ,b)：
Wherein, t rsk For the mean travel time variable of the path rsk,. Kappa.. Is a Gaussian kernel, b is a smoothing parameter called the bandwidth, M denotes the flow direction, M rsk Set of flow directions contained for path rsk;
the smoothing parameter b takes the following values:
wherein the content of the first and second substances,
is the sample variance used to perform KDE;
as shown in fig. 1, in the method, the estimation of the travel time distribution of the path unit is completed based on formulas (1) to (3), and the travel time distribution of any path of the network is estimated by formulas (4) to (8), which are marked as: path travel time distribution estimation (a priori); however, the estimation of the travel time distribution of the path unit and the estimation of the travel time distribution of any path of the road network are both distribution estimation by using the vehicle track data information of the internet, and the consistency of the travel time parameter and the parameter observed by the automatic vehicle identification cannot be ensured. In addition, since the networked vehicles in the urban road network are usually a biased sample, the method in the present application needs to further provide the steps of travel time alignment and adjustment for data fusion. The path travel time alignment model is first derived and then the flow direction travel time is adjusted.
The path travel time alignment model is derived as follows.
External acquisition-based path flow data f rsk (OD versus path flow for path k between rs), the travel time adjuster aims to align the a priori travel time provided by the previously fitted GMM to coincide with the observed travel time between AVI detectors.
Therefore, the present invention takes the formula (9) and the formula (10) as the constraint condition of the optimization formula (11). Where equation (9) describes that the average traveltime recorded between any pair of AVI detectors is constantly equal to the sum of the traveltimes of the paths weighted by that AVI detector, and equation (10) describes that the local path traveltime is constantly equal to the sum of the flow direction traveltimes on that local path. The formula (11) is constrained by the system equation formula (9) and the formula (10), so that the situation that the initial travel time distribution obtained based on the internet vehicle estimation is biased can be guaranteed to be still consistent with the AVI. Meanwhile, the travel time information of AVI local observation and the global distribution information provided by the Internet vehicle are fully combined, and the problem that the travel time of the whole road network is difficult to estimate by a single AVI data source is solved.
The constraint conditions are as follows:
wherein, t p Represents the average of the recorded travel times between any pair of AVI detectors;
representing a travel time between a target pair of AVIs along path rsk; RS is a set of OD pairs, K rs Is the set of paths between the OD versus the rs,
indicates whether the local path p is passed by the path rsk;
is the occurrence indicating whether the flow direction m is passed by rsk and is within the range of the AVI sequence p; f. of rsk Representing path traffic data;
the residual term is defined as the residual time of the path rsk passing through the AVI sequence p, which is the basis for adjusting the flow direction run time.
S12: constructing an optimized objective function based on the least square error based on AVI data:
wherein P represents a local path, P is a set of paths, RS is a set of OD pairs of traffic start and stop points, K rs Is the path set between OD to rs; z is the function of the object function,
in the objective function, the first term
Is a typical OLS entry;
item II
Is to minimize the variance of the residual error for different paths of the same flow direction combination between p;
w' 1 ,w' 2 are the weights of the first and second terms, respectively, and v is the standard deviation of the vehicle travel time between the recorded pair of AVI data p. The default value of the ratio of the two weights is w' 1 :w' 2 1, in specific implementation, w 'in this embodiment may be set according to the confidence level of two targets in actual operation' 1 ,w' 2 The optimal weight ratio of (1): 100.
the optimization target constructed by the formula (11) can be solved by a basic quadratic programming solving method, and a solution with the minimum travel time mean difference between the same AVI pair in different paths is innovatively found out under the condition that the observed quantity of the average travel time of multi-source data (networked vehicle trajectory data and AVI data) is consistent, namely an optimal solution with common compatibility of the multi-source data.
In specific implementation, the optimization model MinZ is standard convex quadratic programming, and can be directly solved by adopting classical methods such as an interior point method and the like, or an external solver is called for solving. Wherein t is p And v are obtained by counting the mean and standard deviation of the differences between matching vehicle timestamps between AVI pairs p respectively,
based on the road network topology (whether the path passes through two sections in AVI pair p), f rsk By external input (path flow estimate or traffic survey statistics).
The flow direction travel time is adjusted as follows.
S13: based on constraint conditions, solving the objective function minimum objective function minZ to obtain decision variables
The solution of (1);
the following formula is adopted for decision variables
The solution of (c) is converted into an adjustment of the average flow direction travel time;
wherein the content of the first and second substances,
indicating the adjustment ratio allocated to the flow direction m on the path rsk in the range of AVI to p; t is t m Is the travel time in the flow direction m;
represents the average travel time adjustment amount of the flow direction m;
indicating the occurrence of whether the flow direction m is passed by rsk and within the scope of the AVI sequence p; m is a set of all flow directions included in the road network to be analyzed;
s14: assigning adjustments to each component of the GMM containing the target flow direction according to the constituent weights:
wherein c is the component of GMM, mu c Is the average value of the component distribution c, lambda c Is the weight of component c.
In the method, the adjustment quantity of a local path p is distributed to each flow direction forming the local path according to a physical space mapping relation through a formula (12) and a formula (13), so that the flow direction adjustment quantity is obtained, and the new flow direction travel time is calculated; after obtaining the new flow direction travel time, the mean value of the GMM is also adjusted correspondingly by the formula (14), i.e. the step of "path-flow direction modulation amount allocation" in fig. 1.
Namely, in the method, accurate identification results are obtained after complementation of the online vehicle sample data covered by the space in full and the AVI data covered by the space in partial.
S15: finally, based on the adjusted GMM, t is recalculated rsk Probability density function tau (t) rsk And b), obtaining the average travel time of the target path rsk.
According to the method for estimating the travel time distribution of the urban road network, firstly, two-dimensional travel time distribution of the road units is obtained through Gaussian mixture model fitting based on the track data of the networked vehicles of each road unit in the road network; then, based on the travel time distribution of each path unit, a travel time sample set of any path in the road network can be constructed by a Monte Carlo sampling method, and the probability density of the travel time distribution of the path is generated by a kernel density estimation method; and finally, aligning and adjusting the route travel time distribution by combining the average travel time from the point to the point obtained by the vehicle automatic identification data, and finally, distributing the travel time of all the flow directions and any route in the data network. Compared with the prior art, the method has the advantages of fully fusing and mining the existing observation data, more detailed travel time distribution modeling unit, easy acquisition of fused input data, low cost, strong use flexibility and the like.
The calculation procedure of the method of the present application is described below as example 1.
As shown in table 1, is an example of AVI data acquired.
Table 1: embodiments of AVI data
Table 2 is an example of vehicle trajectory data for the internet connected vehicle.
Table 2: embodiment of vehicle track data of internet vehicle
| Vehicle ID | Time stamp | Road section | Distance traveled on road |
| #V558 | 2022/02/02 10:31:20 | 112 | 35.6 |
| #V558 | 2022/02/02 10:31:23 | 112 | 62.1 |
Based on the data in tables 1 and 2, the path units (intersection 1-north straight, intersection 2-north left turn) GMM are fitted, wherein the component score C is 2:
brought into
In (1), obtaining:
0.4*τ 1 (t - ,t + |[26.2,30.5],[7.3,3.4] 2 )+0.6*τ 2 (t - ,t + |[28.5,32.2],[3.1,0.7] 2 )
the average straight travel time of the 1-north line flowing to the intersection is 27.4 seconds;
example of path traffic (OD vs r =1,s =7, total 7 alternative paths):
f rsk ＝[5,12,6,74,32,2,68]
the optimization model outputs each path flow adjustment quantity example:
example of adjustment amount of straight-going average travel time flowing to 1-north of intersection:
the average travel time of straight going to 1-north of the intersection is 27.4+0.7=28.1 seconds
The GMM after adjustment is:
0.4*τ 1 (t - ,t + |[26.5,30.5],[7.3,3.4] 2 )+0.6*τ 2 (t - ,t + |[28.9,32.2],[3.1,0.7] 2 )
sample 3 samples example from this GMM:
[(27.3,30.7),(29.5,32.1),(24.1,34.5)]
OD versus 3 sample examples of path travel time of r =1,s =7,k =1:
[170.5,186.2,199.2]
final output OD versus path travel time profile of r =1,s =7,k =1:
next, a simulation case is taken as an embodiment 2, and the travel time distribution estimation method of the present application is verified, as shown in fig. 3, the road network is located in the south of a certain city and is composed of 25 intersections, of which 18 are signalized intersections (including 14 four-arm intersections and 4T-shaped intersections). The invention establishes a simulation model by using VISUM and VISSIM to comprehensively evaluate the estimation effect of travel time distribution. The simulation model was calibrated in the tflowwuzzy module of VISUM using traffic collected from the electric police checkpoint, travel time collected from local taxis, and signal timing schemes collected from local law enforcement. All data were collected in 2019 at month 3.
Figure 4 illustrates the AVI deployment location. According to the calibration result, the absolute error of the flow is less than 6veh/h, and the relative error is less than 9%. In the calibrated network, there are 28 OD nodes and 305 paths. To generate the required data sources, some vehicles are randomly sampled as networked vehicles. The simulation model runs for a total of two (simulation) hours and outputs the required data file.
All the following experiments were performed on a server with a 2.8GHz six-core CPU and 8GB memory. The method adopts the evaluation indexes of Mean Absolute Error (MAE) and mean percent error (MAPE) to evaluate the estimation effect of the method.
In this embodiment, a bimodal GMM (i.e., a component number of 2) is used to fit the flow-to-travel-time distribution, which therefore divides the path-unit distribution into four modes (slow-slow, slow-fast, fast-slow, fast-fast). In this case, the number of components of the GMM-fitted path per travel time is set to 4, and the path travel time distribution is approximated from a distribution of 2,000 samples generated by the KDE. The evaluation results are shown in table 3 below, and it can be seen that the road network travel time estimation method provided by the present application can achieve better travel time distribution estimation both at the flow direction and at the path level.
Table 3: evaluation results
Because the problems of low uploading frequency, low market permeability, biased samples and the like exist in the floating car, the stability of estimation and the reliability of the estimated travel time distribution cannot be guaranteed; meanwhile, the distribution density of the AVI data represented by the electric police access in the urban road network is very limited, and the universality of the travel time distribution estimation in the road network cannot be guaranteed. In the technical scheme, after the path rsk is obtained based on the track data of the floating car, the travel time distribution probability density of any path is calculated, and then the time is calibrated through AVI data.
Claims (4)
1. A method for estimating the travel time distribution of an urban road network based on data fusion is characterized by comprising the following steps:
s1: determining a road network to be analyzed;
s2: acquiring a road network graph of the road network to be analyzed, wherein the road network graph comprises the flow directions of all roads;
s3: constructing a path unit travel time distribution function;
the path unit is as follows: defining two continuous flow directions in a road network as a path unit;
based on the travel time of the network connection vehicle track data in the road network to be analyzed, fitting initial path unit travel time distribution by using a bivariate Gaussian mixture model GMM, wherein the PDF form of the initial path unit travel time distribution function is given by the following formula:
wherein, t - And t + Representing the travel time of the upstream and downstream flow directions, respectively; c represents the number of component distributions in the GMM, λ c Is the weight of component c; tau is c A bivariate gaussian distribution probability density function representing the component distribution c; mu.s c And
is the mean and variance of the component distribution c; τ () represents a bivariate gaussian distribution probability density function of the population;
s4: fitting the initial path unit travel time distribution function GMM through a maximum likelihood function L based on the network connection vehicle track data of each path unit in the road network;
wherein E is a path unit set of all paths included in the road network to be analyzed; n is e The path unit E belongs to the number of the network connection vehicle samples observed on the E; t is t i- And t i+ Representing travel times of the ith networked vehicle in the upstream and downstream flow directions;
s5: calculating the average travel time t of each flow direction m in the road network to be analyzed m ：
Wherein, the first and the second end of the pipe are connected with each other,
mapping relationship between path unit e and flow direction m, if path unit e includes flow direction m
Is 1, otherwise
Is 0;
s6: acquiring a target path rsk;
setting: the number of path units included in the target path rsk is NUM;
s7: distributing GMM tau from initial path unit travel time 0 (t-,t + ) In the random generation of N samples
Initializing a variable nu to 0;
s8: make it
And distributed from the condition
Sampling;
s9: let nu = nu +1;
s10: circularly executing the steps S8 to S9 until all NUM path units in the target path rsk are listed;
s11: t obtained by approximating sample distribution by using kernel density estimation KDE method rsk Probability density function tau (t) rsk ,b)：
Wherein, t rsk For the mean travel time variable of the path rsk,. Kappa.. Cndot.is a Gaussian kernel, b is a smoothing parameter called the bandwidth, M denotes the flow direction, M rsk Set flow directions contained for path rsk;
s12: constructing an objective function based on AVI data:
wherein P represents a local path, P is a set of paths, RS is a set of OD pairs of traffic start and stop points, K rs Is the path set between OD to rs; z is the function of the object function,
in the objective function, the first term
Is a typical OLS entry;
second item
Is to minimize the variance of the residual error between p for different paths of the same flow direction combination;
w’ 1 ,w' 2 the weights of the first term and the second term, respectively, v is the standard deviation of the vehicle travel time between the recorded pair of AVI data p;
s13: solving a minimum objective function minZ based on constraint conditions to obtain a decision variable
The solution of (2);
the following formula is used to determine the variables
The solution of (c) is converted into an adjustment of the average flow direction travel time;
wherein the content of the first and second substances,
indicating the adjustment ratio allocated to the flow direction m on the path rsk in the range of AVI to p; t is t m Is the travel time in the flow direction m;
represents the average travel time adjustment amount of the flow direction m;
indicating the occurrence of whether the flow direction m is passed by rsk and within the scope of the AVI sequence p; m is a set of all flow directions included in the road network to be analyzed;
s14: an adjustment is assigned to each component of the GMM containing the target flow direction according to the constituent weight:
wherein c is the component of GMM, mu c Is the average value of the component distribution c, lambda c Is the weight of component c;
s15: recalculating t based on the adjusted GMM rsk Probability density function tau (t) rsk And b), obtaining the average travel time of the target path rsk.
2. The urban road network travel time distribution estimation method based on data fusion as claimed in claim 1, wherein: the smoothing parameter b takes the following values:
wherein the content of the first and second substances,
is the sample variance used to perform KDE.
3. The urban road network travel time distribution estimation method based on data fusion according to claim 1, characterized in that: the constraint conditions are as follows:
wherein, t p Represents the average of the recorded travel times between any pair of AVI detectors;
representing a travel time between a target pair of AVIs along path rsk; RS is a set of OD pairs, K rs Is the set of paths between the OD versus the rs,
indicates whether the local path p is passed by the path rsk;
is the occurrence indicating whether the flow direction m is passed by rsk and is within the range of the AVI sequence p;
the residual error term is defined as the residual travel time of the path rsk through the AVI sequence p; f. of rsk Representing path traffic data.
4. A substrate according to claim 1The method for estimating the travel time distribution of the urban road network based on data fusion is characterized by comprising the following steps: in step S12, the default value of the ratio of the two weights is w' 1 :w' 2 ＝1:1。
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211368233.8A CN115798198B (en) | 2022-11-03 | 2022-11-03 | Urban road network travel time distribution estimation method based on data fusion |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211368233.8A CN115798198B (en) | 2022-11-03 | 2022-11-03 | Urban road network travel time distribution estimation method based on data fusion |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115798198A true CN115798198A (en) | 2023-03-14 |
| CN115798198B CN115798198B (en) | 2024-04-05 |
Family
ID=85435163
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211368233.8A Active CN115798198B (en) | 2022-11-03 | 2022-11-03 | Urban road network travel time distribution estimation method based on data fusion |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115798198B (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103440764A (en) * | 2013-08-19 | 2013-12-11 | 同济大学 | Urban road network vehicle travel path reconstruction method based on vehicle automatic identification data |
| WO2018122585A1 (en) * | 2016-12-30 | 2018-07-05 | 同济大学 | Method for urban road traffic incident detecting based on floating-car data |
| CN110634285A (en) * | 2019-08-05 | 2019-12-31 | 江苏大学 | Road section travel time prediction method based on Gaussian mixture model |
| WO2021109318A1 (en) * | 2019-12-03 | 2021-06-10 | 东南大学 | Method for estimating and predicting short-term traffic circulation state of urban road network |
| CN113129605A (en) * | 2021-03-24 | 2021-07-16 | 同济大学 | Electronic police data-based intersection lane queuing length estimation method |
-
2022
- 2022-11-03 CN CN202211368233.8A patent/CN115798198B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103440764A (en) * | 2013-08-19 | 2013-12-11 | 同济大学 | Urban road network vehicle travel path reconstruction method based on vehicle automatic identification data |
| WO2018122585A1 (en) * | 2016-12-30 | 2018-07-05 | 同济大学 | Method for urban road traffic incident detecting based on floating-car data |
| CN110634285A (en) * | 2019-08-05 | 2019-12-31 | 江苏大学 | Road section travel time prediction method based on Gaussian mixture model |
| WO2021109318A1 (en) * | 2019-12-03 | 2021-06-10 | 东南大学 | Method for estimating and predicting short-term traffic circulation state of urban road network |
| CN113129605A (en) * | 2021-03-24 | 2021-07-16 | 同济大学 | Electronic police data-based intersection lane queuing length estimation method |
Non-Patent Citations (3)
| Title |
|---|
| 刘磊等: "基于决策树模型的信号控制交叉口交通状态估计", 《公路交通科技》, 30 September 2019 (2019-09-30) * |
| 唐克双等: "An entendable gaussian mixture model for lane-based queue length estimation based on license plate recognition data", 《JOURNAL OF ADVANCED TRANSPORTATION》, 29 December 2022 (2022-12-29) * |
| 李瑞敏;钱小冬;武红斌;: "城市道路旅行时间高斯混合模型研究", 交通运输系统工程与信息, no. 04, 15 August 2016 (2016-08-15) * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115798198B (en) | 2024-04-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Seo et al. | Estimation of flow and density using probe vehicles with spacing measurement equipment | |
| WO2018064931A1 (en) | Method for estimating travel time distribution of taxi on urban roads when operating states of taxis are considered | |
| Zhan et al. | Urban link travel time estimation using large-scale taxi data with partial information | |
| CN109035784B (en) | Dynamic traffic flow OD estimation method based on multi-source heterogeneous data | |
| EP3753002B1 (en) | Methods and systems for generating traffic volume or traffic density data | |
| US20180286224A1 (en) | System and method of traffic survey, traffic signal retiming and traffic control | |
| CN114239371A (en) | Simulation-based parameter calibration method for vehicle delay model at entrance and exit of large parking lot | |
| Baek et al. | Accurate vehicle position estimation using a Kalman filter and neural network-based approach | |
| CN105869402A (en) | Highway section speed correction method based on multiple types of floating car data | |
| Dantsuji et al. | A novel metamodel-based framework for large-scale dynamic origin–destination demand calibration | |
| Zhang et al. | Extracting origin-destination with vehicle trajectory data and applying to coordinated ramp metering | |
| Nguyen et al. | DFROUTER—Estimation of vehicle routes from cross-section measurements | |
| CN110675631A (en) | Traffic flow traceability analysis method and system | |
| Alrukaibi et al. | Real-time travel time estimation in partial network coverage: A case study in Kuwait City | |
| Park et al. | Model for filtering the outliers in DSRC travel time data on interrupted traffic flow sections | |
| CN115798198B (en) | Urban road network travel time distribution estimation method based on data fusion | |
| Gong et al. | Estimating link travel time with sparse GPS data on highway corridors | |
| CN107886192B (en) | Data and information fusion method based on fixed and mobile vehicle detection data | |
| Fortuijn et al. | Capacity estimation on turboroundabouts with gap acceptance and flow level methods | |
| CN113421428A (en) | Travel impedance model calibration and check method based on multi-source GPS data | |
| Bauer et al. | Modelling travel time uncertainty in urban networks based on floating taxi data | |
| CN116129631A (en) | Vehicle-road cooperative data processing method and related equipment | |
| Hui et al. | Estimation of time-varying OD demands incorporating FCD and RTMS data | |
| Han et al. | Spatiotemporal congestion recognition index to evaluate performance under oversaturated conditions | |
| El Esawey et al. | Using buses as probes for neighbor links travel time estimation in an urban network |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |