CN111915874A - Road average passing time prediction method - Google Patents
Road average passing time prediction method Download PDFInfo
- Publication number
- CN111915874A CN111915874A CN201910378265.8A CN201910378265A CN111915874A CN 111915874 A CN111915874 A CN 111915874A CN 201910378265 A CN201910378265 A CN 201910378265A CN 111915874 A CN111915874 A CN 111915874A
- Authority
- CN
- China
- Prior art keywords
- road
- data set
- time
- average
- context information
- 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
- 238000000034 method Methods 0.000 title claims abstract description 39
- 238000012549 training Methods 0.000 claims description 11
- 238000010801 machine learning Methods 0.000 claims description 4
- 238000000605 extraction Methods 0.000 description 11
- 238000005516 engineering process Methods 0.000 description 5
- 238000002474 experimental method Methods 0.000 description 4
- 239000000284 extract Substances 0.000 description 4
- 230000000694 effects Effects 0.000 description 3
- 230000004927 fusion Effects 0.000 description 3
- 238000001514 detection method Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 238000013459 approach Methods 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000001186 cumulative effect Effects 0.000 description 1
- 238000007418 data mining Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 230000001939 inductive effect Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 238000011144 upstream manufacturing Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/01—Detecting movement of traffic to be counted or controlled
- G08G1/0104—Measuring and analyzing of parameters relative to traffic conditions
- G08G1/0125—Traffic data processing
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/01—Detecting movement of traffic to be counted or controlled
- G08G1/0104—Measuring and analyzing of parameters relative to traffic conditions
- G08G1/0108—Measuring and analyzing of parameters relative to traffic conditions based on the source of data
Landscapes
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Traffic Control Systems (AREA)
Abstract
The invention discloses a road average passing time prediction method, which comprises the following steps: converting vehicle position information acquired based on a GPS into a road average transit time data set; acquiring a road context information data set according to urban road network information based on the road average traffic time data set; and predicting the average transit time of the road to be estimated by using time interval grouping based on the road average transit time data set and the road context information data set.
Description
Technical Field
The invention relates to computer and data mining, in particular to a road average transit time prediction method based on an urban road network.
Background
Generally, in recent years, introduction of intelligent transportation gradually changes the way people travel. One of its wide applications in modern cities is the estimation of road passage time. In an increasingly complex and enlarged urban road network system, accurately estimating the road traffic time is an important means for guiding people to go out and avoiding potential road congestion.
Road travel time estimation is an important but challenging task in intelligent traffic. The method can well reflect the road congestion condition, and is widely applied to long and short journey time estimation at present. The method has important significance for real-time traffic monitoring, driving direction, route selection and traffic resource scheduling. Existing solutions, such as the support vector regression based transit time prediction (Wu C H, Ho J M, Lee D T. travel-time prediction with support vector regression [ J ]. IEEE Transactions on Intelligent Transportation Systems,2004,5(4):276 @) proposed in 2004 by Chun-Hsin Wu et al are connected to the real-time urban traffic prediction model based dynamic Path inducing System [ J ]. RAEUSIP Journal communication Systems and networks, 2014 1 by Liang, Zilu et al, but not to historical road information, (Wu C H, Ho J M, Lee D. travel-time prediction Systems for dynamic road prediction Systems [ J ]. 2014 85).
The road passage time is affected by many factors, such as the length, width and grade of the road. Generally, it takes a long time to travel on a long, narrow and low-grade road under the same artificial activity and natural conditions. Existing studies use historical data of roads to predict their future transit times, and this approach uses one or more predictive models to fit recent and historical data for a certain road and uses that model to predict the transit time of the road in a short time in the future.
However, the above time estimation methods are all estimation based on the objective attribute of the road itself. Due to human activity, estimates of road passage time tend to be related to time period and date. In modern urban traffic systems, the connection between roads is more and more complex, and whether a road is unobstructed or not is often caused by multiple factors, such as: the number and the congestion degree of the upstream and downstream roads of one road, the adjacent roads and the like. Therefore, when estimating the road passage time, modeling cannot be performed using only the historical data or objective attributes of the road itself. However, the existing estimation method does not utilize this part of information, but models the historical data of the road to be predicted to fit the change rule of the road passing time. The generalization ability of this method is limited because it does not have the ability to generalize to temporary accidents and congestion occurring on adjacent or related roads.
In summary, the conventional road passing time prediction technology does not consider the influence between adjacent roads, so that the method has obvious defects. The existing method can not accurately predict the road passing time under the condition of complex urban road network relation.
The invention provides a method for predicting the average road passing time based on the context information of an urban road network, aiming at solving the problems, the method extracts the context information based on the connection relation of urban roads and combines the extracted information with the average road passing time, so that the extracted context information and the average road passing time have the same dimension, and the accuracy of predicting the average road passing time can be greatly improved. The context information extracted by the invention is completely based on the road passing time acquired by the GPS, and the extra acquisition overhead is not increased; the core algorithm has good expandability and can be completely applied to the current road passing time estimation algorithms of all types; the context information extracted by the extraction technology can greatly improve the accuracy of the road average traffic time estimation; the method can be conveniently applied to urban large-scale roads and has industrial application prospect.
Disclosure of Invention
The invention aims to provide a method for predicting the average road passing time, which can overcome the problems that the existing method has low estimation precision of the road passing time, excessively depends on the historical data of a road to be predicted, has insufficient data of the road to be predicted and the like, and can be effectively applied to the estimation of the average road passing time of a large-scale urban road network.
In order to achieve the above object, according to an aspect of the present invention, there is provided a road mean transit time prediction method, the method including: converting vehicle position information acquired based on a GPS into a road average transit time data set; acquiring a road context information data set according to urban road network information based on the road average traffic time data set; and predicting the average transit time of the road to be estimated by using time interval grouping based on the road average transit time data set and the road context information data set.
The step of converting the vehicle position information obtained based on GPS into a road mean transit time data set includes: the vehicle position information extracted based on the GPS is corresponding to each road in the urban road network, and a GPS positioning sequence of each vehicle on each road is obtained; making difference on adjacent elements of the GPS positioning sequence of each vehicle on each road and dividing the difference by the positioning interval to obtain the passing speed in a plurality of unit time; calculating the average speed of each vehicle speed on each road, and dividing the average speed by the length of the road to obtain the road average passing time of the vehicle on the road; and taking a fixed time length as a time period, counting the road average passing time of all vehicles on each road in the time period, and forming a road average passing time data set by the road average passing time of all vehicles on each road in each time period.
Based on the road average traffic time data set, the step of acquiring a road context information data set according to urban road network information comprises the following steps: acquiring road context information according to an urban road network; and fusing the acquired road context information to acquire a road context information data set.
The step of obtaining the road context information according to the urban road network comprises the following steps: inquiring and traversing each road r in all time intervals according to the urban road network informationi，riIs λiInitial lambda ofi1 is ═ 1; inquiring and traversing the current road r according to the urban road network informationiAll the adjacent roads rj(ii) a According to the road riAnd adjacent road rjLength, width, road grade of r, obtain riAnd rjDegree of correlation λ ofij(ii) a Calculating the adjacent road rjCumulative degree of correlation λj＝λij*λi(ii) a Will be adjacent to the road rjAverage road passing time of (a) multiplied by lambdajAs riFrom rjExtracted context information F (r)i,rj) (ii) a And taking the adjacent road as a new current road, and iteratively executing the steps.
The method for fusing the acquired road context information to acquire the road context information data set comprises the following steps: for the road riAll the context information F (r)i,rj) Add up to obtain the road riThe context information of all time periods of the road form a road context information data set.
The step of predicting the average transit time of the link to be estimated using the time period grouping based on the link average transit time data set and the link context information data set includes: the method comprises a training phase and a prediction phase, wherein in the training phase, the road average transit time data set and the road context information data set are divided into 7 sub-data sets from Monday to Sunday respectively; correspondingly combining 7 sub-data sets of the road average traffic time data set and 7 sub-data sets of the road context information data set to obtain a road traffic information data set, wherein the road traffic information data set comprises 7 sub-data sets divided according to Mondays to Sundays; dividing each subdata set of the road traffic information data set into N road traffic information segments with specific lengths at random according to fixed time length, wherein N is a positive integer; using the divided road traffic information segments, and training by using a machine learning algorithm to obtain 7 road average traffic time prediction models; in a prediction stage, acquiring a road traffic information segment of a road to be estimated within the fixed time length from the road traffic information data set; and inputting the road traffic information segment of the road to be estimated into one of the 7 road average traffic time prediction models to predict the road average traffic time.
According to another aspect of the present invention, there is provided a computer readable storage medium storing computer readable instructions, wherein the computer readable instructions when executed realize the above-mentioned road average transit time prediction method.
According to the method, firstly, a graph is built according to urban road basic information, influence factors between adjacent roads are specified, then, the road context information is respectively extracted based on each road, and the influence factors are continuously accumulated in the extraction process, so that the accumulated influence factors are smaller for the roads which are farther away from the road to be extracted. And fusion operation is realized by adding context information of simultaneous dimensionality. Therefore, the storage overhead is greatly saved, and the extracted context information has the same dimension as the original transit time, so that the method has high expandability.
The technical scheme of the invention extracts the context information based on the connection relation of the urban roads, and can greatly improve the accuracy rate of predicting the road traffic time. The context information extracted by the invention is completely based on the road passing time acquired by the GPS, and the extra acquisition overhead is not increased; the core algorithm has good expandability and can be completely applied to the current road passing time estimation algorithms of all types; the context information extracted by the extraction technology can greatly improve the accuracy of the time estimation of the road; the extraction method can be conveniently applied to large-scale roads in cities, and has industrialized application prospect.
The invention designs and directly merges the road context information data set and the road average transit time data set, so that the extracted data can be applied to a road transit time estimation task without any loss by the same dimension merging method, and experiments prove that the technical scheme can greatly improve the transit time prediction accuracy.
The proposed technology for predicting the average traffic time of the road based on the time interval grouping solves the problem of the influence of the time intervals on the traffic time. The way of grouping and predicting the roads in different time periods can effectively realize the accurate detection of the road passing time. Not only can data be easily stored in a distributed mode in a multi-processing system, but also the algorithm can have extremely high parallelism.
The urban road network context information-based road traffic time prediction technology extracts the context information based on the connection relation of urban roads and combines the extracted context information data set and the road average traffic time data set, so that the extracted context information data set and the road average traffic time data set have the same dimension, and the road average traffic time prediction accuracy can be greatly improved. The context information extracted by the invention is completely based on the road passing time acquired by the GPS, and the extra acquisition overhead is not increased; the core algorithm has good expandability and can be completely applied to the current road average traffic time estimation algorithms of all types; the context information extracted by the extraction technology can greatly improve the accuracy of the time estimation of the road; the extraction method can be conveniently applied to large-scale roads in cities, and has industrialized application prospect.
Drawings
FIG. 1 is a flow chart of a method for predicting average transit time of roads based on context information of a city road network according to an embodiment of the present invention;
FIG. 2 is a schematic flow diagram for converting GPS-based vehicle location information into a road mean transit time data set, according to one embodiment of the present invention;
FIG. 3 is a schematic flow chart of obtaining a road context information data set according to urban road network information based on a road mean transit time data set according to an embodiment of the present invention;
FIG. 4 is a schematic flow chart of obtaining road context information according to a city road network according to an embodiment of the present invention;
FIG. 5 is a schematic flow diagram illustrating the fusion of the acquired road context information according to an embodiment of the present invention; and
fig. 6 is a schematic flowchart of road average transit time prediction for a road to be estimated by using time interval grouping based on an acquired road average transit time data set and a road context information data set according to an embodiment of the present invention.
Detailed Description
The following description is made with reference to the accompanying drawings and provided to assist in a comprehensive understanding of exemplary embodiments of the invention as defined by the claims and their equivalents. The disclosure includes various details to aid understanding, but these are considered examples only. Accordingly, those of ordinary skill in the art will recognize that various changes and modifications of the embodiments described herein can be made without departing from the scope and spirit of the present disclosure. In addition, descriptions of well-known functions and constructions may be omitted for clarity and conciseness.
The terms and words used in the following description and claims are not limited to the written meaning, but are used only to enable a clear and consistent understanding of the disclosure. Accordingly, it should be apparent to those skilled in the art that the following description of the exemplary embodiments of the present disclosure is provided for illustration only and not for the purpose of limiting the disclosure as defined by the appended claims and their equivalents.
The invention is described in detail below with reference to the figures and specific embodiments.
Fig. 1 is a flow chart of a method for predicting average transit time of roads based on context information of a city road network according to an embodiment of the present invention.
The position information of the vehicle traveling on the road may be acquired by a positioning system. The positioning system may be any positioning system capable of acquiring vehicle position information, and may be at least one of a Global Positioning System (GPS), a global navigation satellite system (GLONASS), a BeiDou navigation satellite system (BeiDou), Galileo (Galileo), or a european global satellite navigation system, for example. Hereinafter, description will be made by taking a Global Positioning System (GPS) as an example, and of course, any one of the positioning systems may be used.
Referring to fig. 1, the method for predicting the average road transit time based on the context information of the urban road network provided by the invention comprises the following steps:
in step S1, the vehicle location information obtained based on GPS may be converted into a road mean transit time data set.
Specifically, the conversion of GPS-based acquired vehicle location information into a road mean transit time data set is described with reference to fig. 2.
Specifically, as shown in fig. 2, in step S101, vehicle position information extracted based on GPS is mapped to each road in the urban road network, and a GPS positioning sequence of each vehicle on each road is obtained.
Then, proceed to step 102. In step 102, the difference is made between adjacent elements of the GPS positioning sequence of each vehicle on each road and divided by the positioning interval to obtain the passing speed in a plurality of unit times. The positioning interval is the interval between two adjacent elements of the GPS positioning sequence referred to above.
Next, in step 103, an average speed is obtained for each vehicle speed on each road, and the length of the road is divided by the average speed to obtain the road average transit time of the vehicle on the road.
In step S104, the average road transit time of all vehicles on each road in a time period (for example, 10 minutes, although the time period may be arbitrarily set as the case may be) is counted with a fixed time period. And constructing a road average passing time data set by the road average passing time of all vehicles on each road in each time period.
The road mean transit time data set may be stored in the format of table 1 below.
TABLE 1 road mean transit time data set
The link average transit time data set is an important data structure for storing the link average transit time, which records the link average transit time and the corresponding time period for each link, and the data thereof can be directly used for the training of the prediction model M, which will be described in detail later. The storage format of the road average transit time data set shown in table 1 is only one storage example, and those skilled in the art can store in various formats according to the needs.
Next, returning to fig. 1, step S2 is performed in the method for predicting the average road transit time based on the context information of the urban road network. In step S2, a road context information data set is acquired from the urban road network information based on the road mean transit time data set acquired in step S1.
Specifically, a specific implementation manner of acquiring the road context information data set according to the urban road network information based on the road average transit time data set is described with reference to fig. 3.
Specifically, as shown in fig. 3, in step S201, the road context information according to the urban road network is obtained, and the specific implementation steps are described with reference to fig. 4.
Fig. 4 shows a flow chart for obtaining road context information from a city road network. Specifically, as shown in fig. 4, in step S201-1, each road in all time periods is queried and traversed according to the urban road network information, and the currently traversed road may be defined as riAt this time, riInitializing a road cumulative association lambdaiAnd can be set to an initial λi＝1。
In step S201-2, the current road r is queried and traversed according to the city road network informationiMay define the currently traversed adjacent link as rj。
In step S201-3, according to the road riTo each adjacent road rjLength, width, road grade of, calculate riAnd rjDegree of correlation λ ofijFor two roads of a certain length, width, road class, the degree of association λijCan be specified manually or can be obtained by experiment.
In step S201-4, each of the adjoining roads r is calculated separatelyjI.e. for each rjCumulative degree of correlation λjIs calculated as lambdaj＝λij*λi。
In step S201-5, each of the adjoining roads rjAverage road passing time of (a) multiplied by lambdajAs riFrom rjThe extracted context information is assumed to beF(ri,rj). The road average transit time is obtained from a set of road average transit time data.
Then, in step S201-6, each of the adjoining roads is regarded as a new current road, i.e., r is taken asjIs regarded as riAnd (5) performing S201-2 to S201-6 in an iteration mode, wherein the iteration frequency is manually specified or can be obtained through experiments.
Next, returning to FIG. 3, at the acquired road context information F (r)i,rj) Then, in step 202, the obtained road context information is fused to obtain a road context information data set, and the specific implementation steps are described with reference to fig. 5.
Fig. 5 shows a process of fusing the acquired road context information according to the urban road network. Specifically, as shown in FIG. 5, in step 202-1, each road r in all time periods is queried and traversed according to the urban road network informationi。
Next, in step 202-2, for the currently traversed road riAll F (r) are addedi,rj) Add up to obtain the road riAnd as a road context information data set.
The road context information data set records the context information of each road in each time period. The specific format of the road context information data set is shown in table 2.
TABLE 2 road context information dataset
The storage format of the road context information data set shown in table 2 is only one storage example, and those skilled in the art may store in various formats as needed.
Firstly, drawing is built according to urban road basic information, influence factors between adjacent roads are specified, then road context information is respectively extracted based on each road, and the influence factors are continuously accumulated in the extraction process, so that the accumulated influence factors are smaller for the roads which are farther away from the road to be extracted. And fusion operation is realized by adding context information of simultaneous dimensionality. Therefore, the storage overhead is greatly saved, and the extracted context information has the same dimension as the original transit time, so that the method has high expandability.
The context information is extracted based on the connection relation of the urban roads, so that the prediction accuracy of the average road passing time can be greatly improved. The context information extracted by the invention is completely based on the average road passing time acquired by the GPS, and the extra acquisition overhead is not increased; the core algorithm has good expandability and can be completely applied to the current road average traffic time estimation algorithms of all types; the context information extracted by the extraction technology can greatly improve the accuracy of the time estimation of the road; the extraction method can be conveniently applied to large-scale roads in cities, and has industrialized application prospect.
Then, returning again to fig. 1, after the road context information data set is acquired through step S2, in step S3, average transit time prediction is performed on the road to be estimated using the time-period grouping, based on the road average transit time data set acquired through step S1 and the road context information data set acquired through step S2.
Specifically, a specific implementation of predicting the average transit time of the road to be estimated using the time-interval grouping based on the acquired road average transit time data set and the road context information data set will be described with reference to fig. 6.
As shown in fig. 6, the travel time prediction for the road to be estimated using the time slot grouping is divided into two stages. The first phase is the model training phase. Specifically, in the model training phase, the following steps are performed.
In step S301-1, the road average transit time data set acquired in step S1 and the road context information data set acquired in step S2 are divided into 7 sub-data sets from monday to sunday, respectively, that is, 7 sub-data sets of the road average transit time data set and 7 sub-data sets of the road context information data set are obtained.
In step S301-2, the 7 sub data sets of the road average transit time data set and the 7 sub data sets of the road context information data set obtained by dividing in step S301-1 are directly merged into a road transit information data set including the 7 sub data sets. That is, with each day of the week as a merging criterion, data of monday of the two data sets are merged, data of tuesday are merged, and so on, a road traffic information data set including 7 sub-data sets is obtained. Modeling each day of the week separately takes into account the effect of different dates on the average transit time of the road.
In step S301-3, each sub data set is modeled separately based on the road traffic information data set composed of 7 sub data sets combined in step S301-2. Specifically, each sub data set is randomly divided into N (N is a positive integer) road traffic information segments with specific lengths according to a fixed time length X (the time length can be arbitrarily set according to the situation), where the structure of each road traffic information segment is shown in table 3:
TABLE 3 road traffic information segment
Wherein, the selection of the time hh mm ss is random selection, and the time can be set arbitrarily according to the situation.
Then, for each subdata set, using the divided road traffic information segments, training (e.g., LSTM) is performed using a machine learning algorithm, thereby obtaining road average traffic time prediction models M1-M7, i.e., one model for each day of the week. Other machine learning algorithm training as disclosed in the art is equally applicable.
Then, a prediction phase is entered. In the prediction phase, assuming that the average transit time of the link r after the time HH MM SS is to be predicted, the following steps are performed:
in step S302-1, a road traffic information piece of the road within the time period [ 'HH: MM: SS-X', 'HH: MM: SS' ] is acquired in the road traffic information data set for the road r to be estimated. The structure of the acquired road traffic information segment is shown in table 4:
TABLE 4 road traffic information segment for road r
In step S302-2, the road traffic information segment of the road r to be estimated is input to one of the road average traffic time prediction models M1-M7, and the road average traffic time is predicted.
The invention designs and directly merges the road context information data set and the road average transit time data set, so that the extracted data can be applied to a road transit time estimation task without any loss by the same dimension merging method, and experiments prove that the technical scheme can greatly improve the transit time prediction accuracy.
The proposed technology for predicting the average traffic time of the road based on the time interval grouping solves the problem of the influence of the time intervals on the traffic time. The way of grouping and predicting the roads in different time periods can effectively realize the accurate detection of the road passing time. Not only can data be easily stored in a distributed mode in a multi-processing system, but also the algorithm can have extremely high parallelism.
The urban road network context information-based road traffic time prediction technology extracts the context information based on the connection relation of urban roads and combines the extracted context information data set and the road average traffic time data set, so that the extracted context information data set and the road average traffic time data set have the same dimensionality, and the road average traffic time prediction accuracy can be greatly improved. The context information extracted by the invention is completely based on the road passing time acquired by the GPS, and the extra acquisition overhead is not increased; the core algorithm has good expandability and can be completely applied to the current road average traffic time estimation algorithms of all types; the context information extracted by the extraction technology can greatly improve the accuracy of the time estimation of the road; the extraction method can be conveniently applied to large-scale roads in cities, and has industrialized application prospect.
Although the inventive concept has been described with reference to exemplary embodiments thereof, it will be apparent to those skilled in the art that various changes and modifications may be made therein without departing from the scope of the inventive concept as set forth in the following claims.
Claims (7)
1. A method of road mean transit time prediction, the method comprising:
converting vehicle position information acquired based on a GPS into a road average transit time data set;
acquiring a road context information data set according to urban road network information based on the road average traffic time data set; and
and predicting the average transit time of the road to be estimated by using time interval grouping based on the road average transit time data set and the road context information data set.
2. The method of predicting road mean transit time of claim 1 wherein the step of converting GPS-based vehicle location information into a road mean transit time data set comprises:
the vehicle position information extracted based on the GPS is corresponding to each road in the urban road network, and a GPS positioning sequence of each vehicle on each road is obtained;
making difference on adjacent elements of the GPS positioning sequence of each vehicle on each road and dividing the difference by the positioning interval to obtain the passing speed in a plurality of unit time;
calculating the average speed of each vehicle speed on each road, and dividing the average speed by the length of the road to obtain the road average passing time of the vehicle on the road; and
taking a fixed time period as a time period, counting the road average passing time of all vehicles on each road in the time period, and forming a road average passing time data set by the road average passing time of all vehicles on each road in each time period.
3. The method of claim 1, wherein the step of obtaining a road context information data set from urban road network information based on the road mean transit time data set comprises:
acquiring road context information according to an urban road network; and
and fusing the acquired road context information to acquire a road context information data set.
4. The method of claim 3, wherein the step of obtaining the road context information according to the urban road network comprises:
inquiring and traversing each road r in all time intervals according to the urban road network informationi，riIs λiInitial lambda ofi＝1；
Inquiring and traversing the current road r according to the urban road network informationiAll the adjacent roads rj；
According to the road riAnd adjacent road rjLength, width, road grade of r, obtain riAnd rjDegree of correlation λ ofij；
Calculating the adjacent road rjCumulative degree of correlation λj＝λij*λi；
Will be adjacent to the road rjAverage road passing time of (a) multiplied by lambdajAs riFrom rjExtracted context information F (r)i,rj)；
And taking the adjacent road as a new current road, and iteratively executing the steps.
5. The method of claim 4, wherein the step of fusing the obtained context information of the road to obtain the context information data set of the road comprises:
for the road riAll the context information F (r)i,rj) Add up to obtain the road riThe context information of all time periods of the road form a road context information data set.
6. The method of claim 1, wherein the step of predicting the average transit time of the link to be estimated using the time-interval groups based on the link average transit time data set and the link context information data set comprises: a training phase and a prediction phase, wherein,
in the training stage, dividing the road average transit time data set and the road context information data set into 7 sub-data sets from Monday to Sunday respectively; correspondingly combining 7 sub-data sets of the road average traffic time data set and 7 sub-data sets of the road context information data set to obtain a road traffic information data set, wherein the road traffic information data set comprises 7 sub-data sets divided according to Mondays to Sundays; dividing each subdata set of the road traffic information data set into N road traffic information segments with specific lengths at random according to fixed time length, wherein N is a positive integer; using the divided road traffic information segments, and training by using a machine learning algorithm to obtain 7 road average traffic time prediction models;
in a prediction stage, acquiring a road traffic information segment of a road to be estimated within the fixed time length from the road traffic information data set; and inputting the road traffic information segment of the road to be estimated into one of the 7 road average traffic time prediction models to predict the road average traffic time.
7. A computer readable storage medium storing computer readable instructions, characterized in that the computer readable instructions, when executed by a processor, implement the road average transit time prediction method of any one of claims 1 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910378265.8A CN111915874B (en) | 2019-05-08 | 2019-05-08 | Road average passing time prediction method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910378265.8A CN111915874B (en) | 2019-05-08 | 2019-05-08 | Road average passing time prediction method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111915874A true CN111915874A (en) | 2020-11-10 |
| CN111915874B CN111915874B (en) | 2021-05-28 |
Family
ID=73241963
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910378265.8A Expired - Fee Related CN111915874B (en) | 2019-05-08 | 2019-05-08 | Road average passing time prediction method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111915874B (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112598912A (en) * | 2020-12-10 | 2021-04-02 | 佳都新太科技股份有限公司 | Bayonet interval acquisition method and device, computer equipment and storage medium |
| CN114822016A (en) * | 2022-03-11 | 2022-07-29 | 上海工程技术大学 | Road section uniform speed estimation system based on GPS data flow |
| CN115440063A (en) * | 2022-09-01 | 2022-12-06 | 的卢技术有限公司 | Traffic signal lamp control method and device, computer equipment and storage medium |
Citations (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP1026649B1 (en) * | 1999-02-06 | 2004-09-15 | DaimlerChrysler AG | Method and device to prepare traffic information |
| US20050143903A1 (en) * | 2003-12-30 | 2005-06-30 | Jin Ho Park | Method for determining traffic conditions |
| JP2006292635A (en) * | 2005-04-13 | 2006-10-26 | Denso Corp | Navigation system for vehicle |
| CN1937001A (en) * | 2006-10-19 | 2007-03-28 | 上海交通大学 | Method for obtaining everage speed of city road section traffic flow |
| CN102332210A (en) * | 2011-08-04 | 2012-01-25 | 东南大学 | Method for extracting real-time urban road traffic flow data based on mobile phone positioning data |
| CN103632540A (en) * | 2012-08-20 | 2014-03-12 | 同济大学 | An urban main road traffic operation information processing method based on floating vehicle data |
| CN104269059A (en) * | 2014-10-15 | 2015-01-07 | 河海大学 | City path travel time forecasting method based on multi-source data fusion |
| CN105513365A (en) * | 2016-02-15 | 2016-04-20 | 邱小平 | Road passing-time estimating system based on isochronal positioning monitoring |
| CN105702036A (en) * | 2016-03-23 | 2016-06-22 | 深圳市金溢科技股份有限公司 | Calculation method, apparatus and system for vehicle driving time |
| CN105894809A (en) * | 2014-12-25 | 2016-08-24 | 杭州远眺科技有限公司 | Sectional type urban road traffic state estimation method |
| CN106202577A (en) * | 2016-08-26 | 2016-12-07 | 成都汉康信息产业有限公司 | A kind of multidimensional data abstracting method based on MDA |
| DE102017112865A1 (en) * | 2016-12-14 | 2018-06-14 | Hyundai Motor Company | Apparatus for automated driving control, system with the same, and method therefor |
| CN108417029A (en) * | 2018-02-11 | 2018-08-17 | 东南大学 | City road network travel time estimation method based on adaptive multitask deep learning |
| CN109064750A (en) * | 2018-09-28 | 2018-12-21 | 深圳大学 | Urban road network traffic estimation method and system |
| CN109670277A (en) * | 2019-02-19 | 2019-04-23 | 南京邮电大学 | A kind of travel time prediction method based on multi-modal data fusion and multiple-model integration |
-
2019
- 2019-05-08 CN CN201910378265.8A patent/CN111915874B/en not_active Expired - Fee Related
Patent Citations (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP1026649B1 (en) * | 1999-02-06 | 2004-09-15 | DaimlerChrysler AG | Method and device to prepare traffic information |
| US20050143903A1 (en) * | 2003-12-30 | 2005-06-30 | Jin Ho Park | Method for determining traffic conditions |
| JP2006292635A (en) * | 2005-04-13 | 2006-10-26 | Denso Corp | Navigation system for vehicle |
| CN1937001A (en) * | 2006-10-19 | 2007-03-28 | 上海交通大学 | Method for obtaining everage speed of city road section traffic flow |
| CN102332210A (en) * | 2011-08-04 | 2012-01-25 | 东南大学 | Method for extracting real-time urban road traffic flow data based on mobile phone positioning data |
| CN103632540A (en) * | 2012-08-20 | 2014-03-12 | 同济大学 | An urban main road traffic operation information processing method based on floating vehicle data |
| CN104269059A (en) * | 2014-10-15 | 2015-01-07 | 河海大学 | City path travel time forecasting method based on multi-source data fusion |
| CN105894809A (en) * | 2014-12-25 | 2016-08-24 | 杭州远眺科技有限公司 | Sectional type urban road traffic state estimation method |
| CN105513365A (en) * | 2016-02-15 | 2016-04-20 | 邱小平 | Road passing-time estimating system based on isochronal positioning monitoring |
| CN105702036A (en) * | 2016-03-23 | 2016-06-22 | 深圳市金溢科技股份有限公司 | Calculation method, apparatus and system for vehicle driving time |
| CN106202577A (en) * | 2016-08-26 | 2016-12-07 | 成都汉康信息产业有限公司 | A kind of multidimensional data abstracting method based on MDA |
| DE102017112865A1 (en) * | 2016-12-14 | 2018-06-14 | Hyundai Motor Company | Apparatus for automated driving control, system with the same, and method therefor |
| CN108417029A (en) * | 2018-02-11 | 2018-08-17 | 东南大学 | City road network travel time estimation method based on adaptive multitask deep learning |
| CN109064750A (en) * | 2018-09-28 | 2018-12-21 | 深圳大学 | Urban road network traffic estimation method and system |
| CN109670277A (en) * | 2019-02-19 | 2019-04-23 | 南京邮电大学 | A kind of travel time prediction method based on multi-modal data fusion and multiple-model integration |
Non-Patent Citations (3)
| Title |
|---|
| XIE XU;VANLINT,HANS等: "A generc data assimilation franework for vehicle trajectory reconstruction on signalized urban arterials using particle filters", 《TRANSPORTATION RESEARCH PART C》 * |
| 伊峰: "基于车载GPS数据的交通流路段平均行程时间估计与预测方法研究", 《中国优秀硕士学位论文全文数据库（工程科技Ⅱ辑）》 * |
| 毕松，车磊等: "城市路网路段行程时间预测研究综述", 《计算机仿真》 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112598912A (en) * | 2020-12-10 | 2021-04-02 | 佳都新太科技股份有限公司 | Bayonet interval acquisition method and device, computer equipment and storage medium |
| CN114822016A (en) * | 2022-03-11 | 2022-07-29 | 上海工程技术大学 | Road section uniform speed estimation system based on GPS data flow |
| CN115440063A (en) * | 2022-09-01 | 2022-12-06 | 的卢技术有限公司 | Traffic signal lamp control method and device, computer equipment and storage medium |
| CN115440063B (en) * | 2022-09-01 | 2023-12-05 | 的卢技术有限公司 | Traffic signal lamp control method, device, computer equipment and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111915874B (en) | 2021-05-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111915874B (en) | Road average passing time prediction method | |
| Ben-Akiva et al. | A dynamic traffic assignment model for highly congested urban networks | |
| Zhan et al. | Urban link travel time estimation using large-scale taxi data with partial information | |
| Bertsimas et al. | Travel time estimation in the age of big data | |
| Hellinga et al. | Decomposing travel times measured by probe-based traffic monitoring systems to individual road segments | |
| US8798897B2 (en) | Real-time traffic analysis through integration of road traffic prediction and traffic microsimulation models | |
| Yuan et al. | T-drive: driving directions based on taxi trajectories | |
| Yuan et al. | Driving with knowledge from the physical world | |
| US8755991B2 (en) | Method and structure for vehicular traffic prediction with link interactions and missing real-time data | |
| Coffey et al. | Time of arrival predictability horizons for public bus routes | |
| CN118053319A (en) | Method and system for generating parking related data | |
| US9754226B2 (en) | Urban computing of route-oriented vehicles | |
| Huang et al. | Context-aware road travel time estimation by coupled tensor decomposition based on trajectory data | |
| US11393334B2 (en) | Method, apparatus, and computer program product for detecting changes in road traffic condition | |
| Zhu et al. | Generating route choice sets with operation information on metro networks | |
| Han et al. | Multi-semantic path representation learning for travel time estimation | |
| Samaranayake et al. | Learning the dependency structure of highway networks for traffic forecast | |
| Jagadeesh et al. | Robust real-time route inference from sparse vehicle position data | |
| Yu et al. | A bus-arrival time prediction model based on historical traffic patterns | |
| Kumar et al. | Analytical approach to identify the optimum inputs for a bus travel time prediction method | |
| Tomaras et al. | Crowd-based ecofriendly trip planning | |
| Tsai et al. | Multivariate long and short term LSTM-based network for traffic forecasting under interference: Experiments during COVID-19 | |
| Iyer et al. | Urban Traffic Congestion Mapping Using Bus Mobility Data. | |
| Li et al. | Personalized travel time prediction using a small number of probe vehicles | |
| CN111243280B (en) | Method and device for calculating indexes of traffic state in real time |
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 | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20210528 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |