CN108877211A

CN108877211A - A kind of link travel speed calculation method based on sparse GPS data

Info

Publication number: CN108877211A
Application number: CN201710360793.1A
Authority: CN
Inventors: 夏莹杰; 蒋萌青; 单振宇
Original assignee: Hangzhou Yuantiao Technology Co Ltd
Current assignee: Hangzhou Yuantiao Technology Co Ltd
Priority date: 2017-05-15
Filing date: 2017-05-15
Publication date: 2018-11-23

Abstract

The present invention relates to a kind of link travel speed calculation methods based on sparse GPS data.It solves existing calculate section travel speed and relies on traffic Floating Car quantity, the problem of inaccuracy is calculated when GPS data is sparse in section.Method includes divide according to user by vehicle GPS data establishing user trajectory point set；For each section of single user's traveling, tracing point nearest from the section on upstream and downstream adjacent segments is added in user trajectory point set；To every section, each user is successively calculated according to course bearing in the travel speed in the section；The average value for calculating all user's travel speeds on section calculates the travel speed in all sections as the travel speed in the section.The present invention overcomes, to the heavy dependence of traffic Floating Car quantity, improve the accuracy of link travel speed calculating in general road average-speed acquisition methods；Simplicity is calculated, arithmetic speed is fast, high reliablity.

Description

A kind of link travel speed calculation method based on sparse GPS data

Technical field

The present invention relates to a kind of intelligent transportation system research fields, more particularly, to a kind of drop evaluated error based on dilute Dredge the link travel speed calculation method of GPS data.

Background technique

It is multiple that intelligent transportation system has been successfully applied to traffic administration, induced travel, signal control, safe driving etc. Field.Average travel time for road sections estimation is one of most important theories problem of intelligent transportation system.GPS(Global Positioning System) the Floating Car information acquiring technology traffic state information dynamic acquisition technology emerging as one, There is many advantages, such as round-the-clock, strong real-time, testing cost are small, data precision is high, acquisition range is wide, and in world wide Positive research and extension application is inside obtained.

Correlative study works：Model using vehicle GPS data estimation bicycle travel speed mainly has rate integrating Model and position interpolation model.The former carries out integral estimation bicycle link travel speed, the latter using GPS instantaneous velocity sequence Using the position of section boundaries on either side GPS positioning point, temporal information, using interpolation method estimation vehicle by section boundary when Between, and then calculate bicycle link travel speed.But the GPS data of high frequency sampling is the premise of two above model, otherwise The average travel speed precision being calculated can be remarkably decreased.

Chen Zhi roc, Liu Yuncai exist within 2007《Method for obtaining average speed of traffic flow of urban road sections》Basis is proposed in patent Link length, average running time and average latency obtain road average speed；Sun Dihua, Liao's filial piety are brave etc. to pass through judgement Whether amount of GPS data is abundant, sufficiently obtains the average speed of traffic flow using different methods with insufficient section to data volume Degree, and consider that signal lamp and bus stop bring influence, the section insufficient for data is using historical data and works as Preceding data, calculate road average-speed；Gorgeous, Chang Ande of Jiang Gui in 2011 etc. exists《Section based on GPS data from taxi is flat Equal velocity estimation model》Different type GPS vehicle operation characteristic is considered in paper, devises two based on GPS data from taxi The improved model for estimating road average-speed, the fusion estimation side of road average-speed is devised based on two improved models Method；Fourth great waves in 2015 are outstanding, history hall is practised《Road average speed estimation method based on GPS data from taxi》It is directed in paper Taxi operation characteristic and its GPS data feature, first obtain the valid data that can be used for road speeds, further according to Position interpolation model improves, and makes it the data suitable for low-frequency sampling.

These above-mentioned methods are when acquisition GPS data is insufficient, often through by historical data or progress interpolation fortune It calculates and estimates current section speed, and then there is a certain error for calculating current time road average-speed when Sparse.

Travel speed is the important parameter in one of data important in floating car data and traffic flow theory research With the basic measurement of traffic circulation.But due to the limitation of the distinctive mobility of Floating Car and Floating Car scale, so that portion There is Sparse Problems for pavement branch sections.The section speed calculation method for inquiring into section Floating Car sparse data has practical meaning Justice, thus necessity invent it is a kind of can based on sparse GPS data to section travel speed carry out calculation method.

Summary of the invention

The present invention mainly solves the prior art and calculates dependence traffic Floating Car quantity to section travel speed, in section The problem of speed estimation error is caused when GPS data is sparse, provide a kind of drop evaluated error based on sparse GPS data Link travel speed calculation method.

Above-mentioned technical problem of the invention is mainly to be addressed by following technical proposals：One kind being based on sparse GPS The link travel speed calculation method of data, includes the following steps：

S1. vehicle GPS data are divided according to user, establishes multiple user trajectory point sets；By GPS data into The matching of walking along the street section is divided into multiple users according to user and gathers, and each user's set establishes the track point set of oneself corresponding road section It closes.

S2. each section travelled for single user, will be nearest from the section on the adjacent segments of upstream and downstream Tracing point is added in user trajectory point set；User's each section track point set is sorted out, is travelled in user every Tracing point nearest from the section on the adjacent segments of upstream and downstream is added to user's section track point set by one section In；The tracing point in every section can be ranked up in set according to the time.

S3. to every section, each user is successively calculated according to course bearing in the travel speed in the section；According to it is upper, The downstream road section tracing point nearest from current road segment there are situation, calculate sole user upstream and downstream section most from current road segment The average value of the travel speed of any point on close tracing point to section, then to the section all the points travel speed average value into Row is average, obtains U.S. user in the travel speed in the section.

S4. the average value for calculating all user's travel speeds on section repeats step as the travel speed in the section 2-4 calculates the travel speed in all sections.

Upstream and downstream section tracing point is added in the calculating of link travel speed by the present invention, effectively overcomes general section To the heavy dependence of traffic Floating Car quantity in method for obtaining average speed, the accuracy of link travel speed calculating is improved.

As a preferred embodiment, the detailed process that user trajectory point set is established in step S1 includes the following steps：

S11. in a timeslice, vehicle GPS data are subjected to path adaptation according to place section；It is wrapped in GPS data Include the i.e. corresponding user of license plate number ID, each license plate ID of corresponding vehicle.Path adaptation is sentenced according to GPS data location information Break section where it.Timeslice is a duration interval, can be 5 minutes or other duration intervals.

S12. GPS data is divided by user by multiple set { A according to license plate number id information₁, A₂, A₃……A_n}；

S13. by each user's set A_nEstablish the track point set A in oneself corresponding section_n={ G_{I, j, k}| i ∈ I, j ∈ J, k ∈ K }, wherein i indicates that user, j indicate the section that user is travelled, and k indicates the track on the running section of user place Point, G_{I, j, k}Indicate that tracing point k, I, J, K of the user i on the j of section are constant.

As a preferred embodiment, tracing point nearest from the section on the adjacent segments of upstream and downstream is added in step S2 Include the following steps to the detailed process in user trajectory set：

S21. the starting point of section j is set as c₁, upstream adjacent segments j-1 is apart from section starting point c₁Nearest point is a,

As any point G on the adjacent segments j-1 of upstream_{I, j-1, k}' meet following formula,

The then G_{I, j-1, k}' it is in upstream adjacent segments j-1 apart from section j starting point c₁Nearest point a, wherein G_{I, j-1, k}' table Show the tracing point on the j-1 of upstream section,Indicate section j starting point c₁To each point G_{I, j-1, k}' distance, | G_{I, j-1, k}′-c₁| it indicates in plane coordinates midpoint G_{I, j-1, k}' arrive starting point c₁Distance；

S22. the terminal of section j is set as c simultaneously₁, downstream adjacent segments j+1 is apart from road segment end c₂Nearest point is b,

As any point G on the adjacent segments j+1 of upstream_{I, j+1, k}' meet following formula,

The then G_{I, j+1, k}' it is in downstream adjacent segments j+1 apart from section j terminal c₂Nearest point b, wherein G_{I, j+1, k}' table Show the tracing point on downstream road section j+1,Indicate section j terminal c₂To each point G_{I, j+1, k}' distance, | G_{I, j+1, k}′-c₂| it indicates in plane coordinates midpoint G_{I, j+1, k}' arrive terminal c₂Distance；

S23. a, b two o'clock are added to user's set A_nTrack point set in { a, G_{I, j, k}, b }.

As a preferred embodiment, it includes following procedure that each user calculates in the travel speed of section j in step S3：

1) in the presence of works as a and b,

If there is no tracing point on section, V_i=S (a-b)/T (a-b), V_iIndicate stroke of the user i on the section Speed, the distance of S (a-b) expression a point to b point, T (a-b) indicate the time used in a point to b point；

If there is tracing point on section, corresponding each tracing point, all calculating tracing point a point to the tracing point, the tracing point To the average value of the travel speed of the travel speed and tracing point a to tracing point b of b point, user link travel speed is then sought Degree,

N indicates the sum of the tracing point on this section；

2) is when a and b only exists 1,

If there are tracing points on the section, for each tracing point, all calculate the tracing point and a point, or with b point The travel speed formed is locked, user's link travel speed is then sought,

When having and in the presence of only a,

When having and in the presence of only b,

3) works as a and b and is not present, and calculates user's link travel there are when 2 tracing points or more on the section Speed,

4) is not present or when a and b only exists 1 as a and b, and the section does not have tracing point or only 1 tracing point When, then user's link travel speed is not calculated.

As a preferred embodiment, the travel speed acquisition process in section includes in step S4：

The average value for calculating the travel speed of all user i in this section, using the average value as the stroke in section speed V is spent,

Repeat the travel speed that step 2-4 calculates all sections.

Therefore, it is an advantage of the invention that：1. upstream and downstream section tracing point is added in the calculating of link travel speed, have Effect overcomes the heavy dependence in general road average-speed acquisition methods to traffic Floating Car quantity, improves link travel The accuracy that speed calculates；2. the travel speed of user can be calculated according to the running data of user whithin a period of time, calculate Simplicity, arithmetic speed is fast, high reliablity, can be effective for road net traffic state assessment, urban traffic control and traffic guidance In.

Detailed description of the invention

Attached drawing 1 is one of present invention flow diagram

Specific embodiment

Below with reference to the embodiments and with reference to the accompanying drawing the technical solutions of the present invention will be further described.

Embodiment：

A kind of link travel speed calculation method based on sparse GPS data of the present embodiment, in order to verify the calculating of proposition Method, the track data that 5 taxi users are chosen in the present embodiment is that research object is tested.This five user's travelings Road sections part it is consistent, track point data is divided into more, general and sparse three classes, representative.The present embodiment is with 5min Counting period.As shown in Figure 1, including the following steps：

S1. vehicle GPS data are divided according to user, establishes multiple user trajectory point sets；

Detailed process includes the following steps：

S11. in a timeslice, vehicle GPS data are subjected to path adaptation according to place section；From database The GPS data of taxi in 5 minutes is read, GPS data essential information includes Customs Assigned Number, tracing point time, tracing point position Coordinate, Customs Assigned Number are associated with license plate number ID.GPS data is subjected to path adaptation according to tracing point position coordinates.

S12. after path adaptation, GPS data is divided by user by multiple set { A according to license plate number id information₁, A₂, A₃……A_n}。

S13. by each user's set A_nEstablish the track point set A in oneself corresponding section_n={ G_{I, j, k}| i ∈ I, j ∈ J, k ∈ K }, wherein i indicates that user, j indicate the section that user is travelled, and k indicates the track on the running section of user place Point, G_{I, j, k}Indicate that tracing point k, I, J, K of the user i on the j of section are constant.Test the user trajectory point set such as table 1 chosen Shown, tracing point sorts in chronological order.

Table 1

S2. each section travelled for single user, will be nearest from the section on the adjacent segments of upstream and downstream Tracing point is added in user trajectory point set, and specific steps include：

The then G_{I, j-1, k}' it is in upstream adjacent segments j-1 apart from section j starting point c₁Nearest point a, wherein G_{I, j-1, k}' table Show the tracing point on the j-1 of upstream section,Indicate section j starting point c₁To each point G_{I, j-1, k}' distance, | G_{I, j-1, k}′- c₁| it indicates in plane coordinates midpoint G_{I, j-1, k}' arrive starting point c₁Distance；

The then G_{I, j+1k}' it is in downstream adjacent segments j+1 apart from section j terminal c₂Nearest point b, wherein G_{I, j+1, k}' indicate Tracing point on downstream road section j+1,Indicate section j terminal c₂To each point G_{I, j+1, k}' distance, | G_{I, j+1, k}′- c₂| it indicates in plane coordinates midpoint G_{I, j+1, k}' arrive terminal c₂Distance；

By taking user 1 as an example, the track point set for listing section 1,2,3,4 is as shown in table 2.

Customs Assigned Number	Section number	Track point set
			User 1	Section 1	{G_1,1,1, G_1,1,2, G_1,1,3, G_{1, Isosorbide-5-Nitrae}, G_1,2,1}
User 1	Section 2	{G_{1, Isosorbide-5-Nitrae}, G_1,2,1, G_1,2,2, G_1,2,3, G_1,2,4, G_1,2,5, G_1,2,6, G_1,3,1}
			User 1	Section 3	{G_1,2,6, G_1,3,1, G_1,3,2, G_1,3,3, G_1,3,4, G_{Isosorbide-5-Nitrae, 1}}
User 1	Section 4	{G_1,3,4, G_{Isosorbide-5-Nitrae, 1}, G_{Isosorbide-5-Nitrae, 2}}

Table 2

S3. to every section, each user is successively calculated according to course bearing in the travel speed in the section；Including with Lower process：

1) in the presence of works as a and b,

If there is no tracing point on section, V_i=S (a-b)/T (a-b), V_iIndicate stroke speed of the user i on the section Degree, the distance of S (a-b) expression a point to b point, T (a-b) indicate the time used in a point to b point；

N indicates the sum of the tracing point on this section；

2) is when a and b only exists 1,

When having and in the presence of only a,

When having and in the presence of only b,

It repeats the process and calculates all users in user's travel speed in each section, by taking 4 users as an example, such as table 3 It is shown.

Customs Assigned Number	1 speed of section	2 speed of section	3 speed of section	4 speed of section
					User 1	17.68km/h	20.13km/h	34.11km/h	29.91km/h
User 2	19.21km/h	25.67km/h	28.76km/h	27.83km/h
					User 3	20.18km/h	18.44km/h	29.87km/h	31.44km/h
User 4	18.32km/h	23.56km/h	30.85km/h	32.31km/h

Table 3

S4. the average value for calculating all user's travel speeds on section repeats step as the travel speed in the section 2-4 calculates the travel speed in all sections.Process includes：

Repeat the travel speed that step 2-4 calculates all sections.Calculated result is as shown in table 4.

Table 4

In addition to the experiment in said one period, we have also carried out peak period and low peak period experiment.In experiment, acquire 50 sections of downtown area the same area same period morning peak data and low peak period data in 30 days, first calculate road Average speed, as right value；Then performance is used for the floating car data calculating speed of the more high-frequency high-precisions of period Estimation.The method that we compare the road speeds method of average and this patent.If calculated result is identical with actual result, indicate to estimate Meter is correct, otherwise estimates mistake.Experimental result is that the accuracy rate of the road speeds method of average is 75.9%, and this patent method is 85.9%.This illustrates that a kind of link travel speed calculation method towards taxi GPS sparse data of the invention is a kind of standard The higher road average-speed calculation method of exactness.

Specific embodiment described herein is only an example for the spirit of the invention.Technology belonging to the present invention The technical staff in field can make various modifications or additions to the described embodiments or by a similar method Substitution, however, it does not deviate from the spirit of the invention or beyond the scope of the appended claims.

Claims

1. a kind of link travel speed calculation method based on sparse GPS data, it is characterised in that：Include the following steps：

S2. each section travelled for single user, by tracing point nearest from the section on the adjacent segments of upstream and downstream It is added in user trajectory point set；

S3. to every section, each user is successively calculated according to course bearing in the travel speed in the section；

S4. the average value for calculating all user's travel speeds on section repeats step 2-4 meter as the travel speed in the section Calculate the travel speed in all sections.

2. a kind of link travel speed calculation method based on sparse GPS data according to claim 1, it is characterized in that step The detailed process that user trajectory point set is established in rapid S1 includes the following steps：

S11. in a timeslice, vehicle GPS data are subjected to path adaptation according to place section；

S12. GPS data is divided by user by multiple set { A according to license plate number id information₁, A₂, A₃......A_n}；

S13. by each user's set A_nEstablish the track point set A in oneself corresponding section_n={ G_{I, j, k}| i ∈ I, j ∈ J, k ∈ K }, wherein i indicates that user, j indicate the section that user is travelled, and k indicates the tracing point on the running section of user place, G_{I, j, k} Indicate that tracing point k, I, J, K of the user i on the j of section are constant.

3. a kind of link travel speed calculation method based on sparse GPS data according to claim 2, it is characterized in that step The detailed process packet being added to tracing point nearest from the section on the adjacent segments of upstream and downstream in rapid S2 in user trajectory set Include following steps：

The then G_{I, j-1, k}' it is in upstream adjacent segments j-1 apart from section j starting point c₁Nearest point a, wherein G_{I, j-1, k}' indicate upstream Tracing point on the j-1 of section,Indicate section j starting point c₁To each point G_{I, j-1, k}' distance, | G_{I, j-1, k}′-c₁| table Show in plane coordinates midpoint G_{I, j-1, k}' arrive starting point c₁Distance；

S22. the terminal of section j is set as c simultaneously₂, downstream adjacent segments j+1 is apart from road segment end c₂Nearest point is b,

The then G_{I, j+1, k}' it is in downstream adjacent segments j+1 apart from section j terminal c₂Nearest point b, wherein G_{I, j+1, k}' indicate downstream Tracing point on the j+1 of section,Indicate section j terminal c₂To each point G_{I, j+1, k}' distance, | G_{I, j+1, k}′-c₂| table Show in plane coordinates midpoint G_{I, j+1, k}' arrive terminal c₂Distance；

4. a kind of link travel speed calculation method based on sparse GPS data according to claim 3, it is characterized in that step It includes following procedure that each user calculates in the travel speed of section j in rapid S3：

1) in the presence of works as a and b,

If there is no tracing point on section, V_i=S (a-b)/T (a-b), V_iIndicate travel speed of the user i on the section, S (a-b) indicate a point arrive b point distance, T (a-b) expression a point to b point used in the time；

If there is tracing point on section, corresponding each tracing point, all calculating tracing point a point to the tracing point, the tracing point to b point Travel speed and tracing point a to tracing point b travel speed average value, then seek the user link travel speed,

N indicates the sum of the tracing point on this section；

2) is when a and b only exists 1,

If there are tracing points on the section, for each tracing point, the tracing point and a point are all calculated, or formed with b point Travel speed, then seek user's link travel speed,

When having and in the presence of only a,

When having and in the presence of only b,

3) works as a and b and is not present, and calculates user's link travel speed there are when 2 tracing points or more on the section,

4) is not present or when a and b only exists 1, and the section does not have tracing point or only 1 tracing point as a and b, then User's link travel speed is not calculated.

5. a kind of link travel speed calculation method based on sparse GPS data according to claim 4, it is characterized in that step The travel speed acquisition process in section includes in rapid S4：

The average value for calculating the travel speed of all user i in this section, using the average value as the travel speed V in the section,

Repeat the travel speed that step 2-4 calculates all sections.