CN108257386A

CN108257386A - Driving trace acquisition methods and device

Info

Publication number: CN108257386A
Application number: CN201611248005.1A
Authority: CN
Inventors: 俞颖晔
Original assignee: Hangzhou Hikvision Digital Technology Co Ltd
Current assignee: Hangzhou Hikvision Digital Technology Co Ltd
Priority date: 2016-12-29
Filing date: 2016-12-29
Publication date: 2018-07-06
Anticipated expiration: 2036-12-29
Also published as: CN108257386B

Abstract

The invention discloses a kind of driving trace acquisition methods and devices, belong to field of computer technology.The method includes：The vehicle for obtaining each bayonet passes through record；Vehicle comprising effective vehicles identifications by recording according to vehicle by the time in record is ranked up, the corresponding history driving trace of each effective vehicles identifications is generated according to ranking results；It identifies the stop section in every history driving trace, which is divided by single driving trace according to the stop section identified, counts the first frequency of identical single driving trace；General driving trace is determined from each single driving trace according to the first frequency of each single driving trace.Solve the problems, such as that the driving trace analysis city road network analysis result accuracy recorded in the relevant technologies according to global positioning system in each car-mounted terminal is low；Achieve the effect that improve city road network precision of analysis.

Description

Driving trace acquisition methods and device

Technical field

The present invention relates to field of computer technology, more particularly to a kind of driving trace acquisition methods and device.

Background technology

The phenomenon that exponentially type increases the quantity of the vehicle travelled in urban road at present, urban traffic congestion is increasingly tight Weight.In order to improve the operational efficiency of city road network, the driving trace based on vehicle is needed to analyze city road network.

In the relevant technologies, generally according to car-mounted terminal global positioning system (Global Positioning System, GPS the driving trace for determining vehicle) is positioned, city road network is analyzed further according to these driving traces.

However, the vehicle travelled in urban road at present is not mounted on car-mounted terminal, analysis city road network institute foundation Driving trace it is limited, the accuracy for leading to the result analyzed city road network is low.

Invention content

It is limited in order to solve to analyze driving trace based on city road network in the prior art, cause to carry out city road network The problem of obtained accuracy of result is low is analyzed, an embodiment of the present invention provides a kind of driving trace acquisition methods and devices. The technical solution is as follows：

In a first aspect, a kind of driving trace acquisition methods are provided, the method includes：The vehicle for obtaining each bayonet passes through Record, the vehicle of the bayonet by record vehicles identifications, the vehicle comprising the vehicle by the bayonet pass through it is described The time of bayonet and the mark of the bayonet；Vehicle comprising effective vehicles identifications is passed through according to vehicle in record by recording Time be ranked up, according to the corresponding history driving trace of each effective vehicles identifications of ranking results generation；Identify that every is gone through The history driving trace is divided into single driving trace by the stop section in history driving trace according to the stop section, Count the first frequency of identical single driving trace；According to the first frequency of each single driving trace from each single driving trace Determine general driving trace；Wherein, the traffic road stopped when section is vehicle generation stop over behavior between adjacent bayonet Section.

Since bayonet is widely distributed in city road network, the vehicle of each bayonet may include by all by record through the card Mouthful vehicle vehicle by record, can reflect vehicle history traveling behavior, thus by vehicle pass through record generate The corresponding history driving trace of each effective vehicles identifications, and after the stop section in identifying every history driving trace, According to the stop section identified by the history driving trace be divided into single driving trace, can obtain in city road network big absolutely The single driving trace of Some vehicles；Solves the traveling recorded in the relevant technologies according to global positioning system in each car-mounted terminal The problem of trajectory analysis city road network analysis result accuracy is low；The effect for improving city road network precision of analysis is reached Fruit.

In addition, general traveling is determined from each single driving trace by the first frequency according to each single driving trace Track, since general driving trace is the track that is frequently occurred in the single driving trace of most vehicles in city road network, When relationship of the general driving trace thus got using aforesaid way between the intensive traffic section is analyzed, what be may be such that divides It is more accurate to analyse result.

It is the track corresponding to driving process of vehicle in addition, since there is no stop sections in single driving trace, Therefore the intensive traffic section corresponding to the general driving trace determined from single driving trace according to frequency is there are incidence relation, There is also incidence relations for the bayonet of these the intensive traffic sections.

Optionally, the stop section in every history driving trace of the identification, including：For the history driving trace In adjacent first bayonet combination two-by-two, obtain first bayonet and combine corresponding vehicle and vehicle is used as by the time in record By first bayonet combine transit time；It determines a period corresponding to the transit time, obtains described the Longest of the one bayonet combination in the period corresponding normal pass duration is passed through duration, and when normal pass is a length of not to be sent out The raw current duration stopped, do not driven over the speed limit；The vehicle is calculated according to the transit time to combine by first bayonet Current duration, when the current duration passes through duration more than the longest, by first bayonet combine between section It is determined as the stop section in the history driving trace.

Vehicle is had differences with a road section in the traffic of different periods do not exceeding the speed limit, not in the case of stop over behavior By normal pass duration used in the section there is also difference, two combined by determining vehicle by first bayonet The transit time corresponding period, obtain longest of first bayonet combination in the period corresponding normal pass duration it is current when It is long, which is relatively determined into first bayonet by the current duration that first bayonet combines compared with the longest passes through duration Whether the section between combination is to stop section, improves the accuracy for stopping section identification.

Optionally, the longest for obtaining the first bayonet combination in the period corresponding normal pass duration is led to Row duration, including：Obtain the current duration that each vehicle is combined within the period by first bayonet；It determines described every Duration section where the current duration of a vehicle, statistics are tied positioned at the quantity of the current duration in each duration section according to statistics Fruit generating probability density function；Hypothetical inspection is carried out to the probability density function, the current duration value of hypothesis will be met In maximum value be determined as the first candidate longest and pass through duration；Utilize accumulation and (cumulative sum, CUSUM) algorithm and institute It states probability density function and determines abnormal current duration, the maximum value in the current duration except the current duration of the exception is true It is set to the second candidate longest to pass through duration；Described first candidate longest is passed through into duration and the second candidate longest is passed through in duration Maximum value is determined as longest of first bayonet combination in the period corresponding normal pass duration and passes through duration；Its In, the independent variable of the probability density function is current duration.

Meet the current duration of hypothesis by obtaining and the current duration of exception determined using CUSUM algorithms except it is logical Row duration may filter that in overspeed of vehicle traveling or vehicle way and be used in the case of stop over by section between first bayonet combination Current duration obtain normal pass duration.

Optionally, it is described that the history driving trace is divided by single driving trace according to the stop section, including： The vehicle of sequence first in the history driving trace is determined as a starting point by record, sorts and passes through in last vehicle Record is determined as a terminal；The preceding vehicle that sorts in the stop section is determined as a upper single by record to travel The terminal of track, the posterior vehicle that sorts are determined as the starting point of next single driving trace by record.

Optionally, the method further includes：Extract the second bayonet group to conform to a predetermined condition in every general driving trace It closes；Calculate the support and at least two confidence levels of each second bayonet combination；Support is met into support threshold and at least The second bayonet that one confidence level meets confidence threshold value combines the bayonet combination for being determined as strong incidence relation；Wherein, it is described pre- Whether fixed condition includes presets position of the bayonet in bayonet item number and/or the combination of the second bayonet in the general driving trace Continuously.

Since the bayonet extracted from the general driving trace of same is there are incidence relation, by extracting every general row The the second bayonet combination to conform to a predetermined condition in track is sailed, the different types of bayonet that is mutually related can be excavated, for example, bayonet item Number is associated with bayonet for 2 binomial, and bayonet item number is more than 2 bi directional association bayonet；For another example, position is continuous in general driving trace Association bayonet, the discontinuous association bayonet in position in general driving trace.

The support and at least two confidence levels combined by calculating each second bayonet determines each second bayonet combination Correlation degree between inner bayonet meets support threshold in the support of certain second bayonet combination and at least one confidence level meets During confidence threshold value, it is believed that the correlation degree between second bayonet combination inner bayonet is stronger, and second bayonet combination is determined as The bayonet combination of strong incidence relation can filter out the strong bayonet group of correlation degree between bayonet from the combination of all second bayonets It closes.

Optionally, the support and at least two confidence levels for calculating the combination of each second bayonet, including：For any Kind the second bayonet combination counts the second frequency of the second bayonet combination；To second frequency and general driving trace Quantity asks quotient, the support that obtained result is combined as second bayonet；For appointing in second bayonet combination One bayonet counts the quantity for including other bayonets in addition to any bayonet in the combination of all second bayonets, to described the Two frequencies and statistical result ask quotient, are put using obtained result as any bayonet is corresponding in second bayonet combination Reliability.

Second aspect, provides a kind of driving trace acquisition device, and described device includes：Acquisition module, it is each for obtaining The vehicle of bayonet is by record, and the vehicle of the bayonet is by recording the vehicles identifications comprising the vehicle by the bayonet, institute It states vehicle and passes through the time of the bayonet and the mark of the bayonet；Generation module, for get the acquisition module Vehicle comprising effective vehicles identifications is ranked up by recording according to vehicle by the time in record, is given birth to according to ranking results Into the corresponding history driving trace of each effective vehicles identifications；Identification module, for identifying every of the generation module generation Stop section in history driving trace；Division module, for the stop section identified according to the identification module will described in History driving trace is divided into single driving trace, counts the first frequency of identical single driving trace；First determining module is used In the first frequency according to each single driving trace general driving trace is determined from each single driving trace；Wherein, it is described The intensive traffic section when stopping section as vehicle generation stop over behavior between adjacent bayonet.

Optionally, the identification module, including：First acquisition unit, for in the history driving trace two-by-two Adjacent the first bayonet combination, the corresponding vehicle of acquisition the first bayonet combination are used as vehicle by the time in record and are passed through The transit time of the first bayonet combination；Second acquisition unit, for determining that the first acquisition unit is got current A period corresponding to time obtains first bayonet and combines the longest in the period corresponding normal pass duration Current duration, a length of current duration for not occurring to stop, do not drive over the speed limit during the normal pass；Determination unit, according to described The transit time that first acquisition unit is got calculates the current duration that the vehicle is combined by first bayonet, described Current duration be more than the longest that the second acquisition unit is got pass through duration when, the road between first bayonet is combined Section is determined as the stop section in the history driving trace.

Optionally, the second acquisition unit, is used for：Each vehicle is obtained within the period by first bayonet The current duration of combination；Determine the duration section where the current duration of each vehicle, statistics is positioned at each duration section The quantity of current duration, according to statistical result generating probability density function；Hypothetical inspection is carried out to the probability density function, The maximum value met in the current duration value of hypothesis is determined as the first candidate longest to pass through duration；Using CUSUM algorithms and The probability density function determines abnormal current duration, by the maximum value in the current duration except the current duration of the exception It is determined as the second candidate longest to pass through duration；Described first candidate longest is passed through into duration and the second candidate longest is passed through in duration Maximum value be determined as the longest that first bayonet is combined in the period corresponding normal pass duration and pass through duration；Its In, the independent variable of the probability density function is current duration.

Optionally, the division module, is used for：The vehicle of sequence first in the history driving trace is true by recording It is set to a starting point, sorts and one terminal is determined as by record in last vehicle；By sequence in the stop section preceding Vehicle the terminal of a upper single driving trace is determined as by record, the posterior vehicle that sorts is determined as next by record The starting point of single driving trace.

Optionally, described device further includes：Extraction module conforms to a predetermined condition for extracting in every general driving trace The second bayonet combination；Computing module, for calculate the support of each second bayonet combination of extraction module extraction and At least two confidence levels；Second determining module, for by support meet support threshold and at least one confidence level satisfaction put The second bayonet combination of confidence threshold is determined as the bayonet combination of strong incidence relation；Wherein, the predetermined condition includes default card Whether position of the bayonet in the general driving trace in mouth item number and/or the combination of the second bayonet be continuous.

Optionally, the computing module, including：Statistic unit, for being combined for any second bayonet, described in statistics Second frequency of the second bayonet combination；First computing unit, for the second frequency that the statistic unit counts on it is general The quantity of driving trace asks quotient, the support that obtained result is combined as second bayonet；Second computing unit, is used for For any bayonet in second bayonet combination, count and included in addition to any bayonet in all second bayonet combinations Other bayonets quantity, the second frequency counted on to the statistic unit asks quotient with statistical result, obtained result made For any bayonet second bayonet combination in corresponding confidence level.

Description of the drawings

To describe the technical solutions in the embodiments of the present invention more clearly, make required in being described below to embodiment Attached drawing is briefly described, it should be apparent that, the accompanying drawings in the following description is only some embodiments of the present invention, for For those of ordinary skill in the art, without creative efforts, other are can also be obtained according to these attached drawings Attached drawing.

Fig. 1 is the flow chart of driving trace acquisition methods provided by one embodiment of the present invention；

Fig. 2A is the flow chart for the driving trace acquisition methods that another embodiment of the present invention provides；

Fig. 2 B are the flows in the stop section in the one history traveling rail of identification provided in another embodiment of the present invention Figure；

Fig. 2 C be provided in another embodiment of the present invention according to stop section one history driving trace is divided into The schematic diagram of single driving trace；

Fig. 2 D are that the support and at least two of the second bayonet of calculating combination provided in another embodiment of the present invention is put The flow chart of reliability；

Fig. 3 A are that the first bayonet of acquisition provided in another embodiment of the present invention is combined in certain period corresponding positive normal open Longest in row duration is passed through the flow chart of duration；

Fig. 3 B are the schematic diagrames to the hypothetical inspection of probability density function progress provided in another embodiment of the present invention；

Fig. 4 A are the block diagrams of driving trace acquisition device provided in one embodiment of the invention；

Fig. 4 B are the block diagrams of driving trace acquisition device provided in another embodiment of the present invention.

Specific embodiment

To make the object, technical solutions and advantages of the present invention clearer, below in conjunction with attached drawing to embodiment party of the present invention Formula is described in further detail.

Before the embodiment of the present invention is described in detail, first to some concepts involved in the embodiment of the present invention Or functions of the equipments progress is as described below：

1st, bayonet refers to be provided with the entrance of monitoring device.

2nd, the monitoring device for being arranged on bayonet has following function：The image for the vehicle that shooting passes through the bayonet, by the figure The time that the shooting time of picture passes through the bayonet as the vehicle；Identify image in vehicle vehicles identifications, by the vehicles identifications, The mark of the time and the bayonet passes through record as a vehicle；The vehicle generated is sent by record by network To server.

It is illustrated using vehicles identifications as license plate number.Monitoring device is when having detected that vehicle passes through from bayonet A, to warp The vehicle crossed, which capture, obtains image, and the records photographing time is 9 days 18 November in 2016:09:58.Monitoring device is to capturing To image carry out license plate number and automatically identify, obtain license plate number as capital * * * * *, then a vehicle of monitoring device generation leads to Overwriting can be：Bayonet A_2016-11-918:09:58_ vehicles capital * * * * *.

3rd, server has following function：Vehicle is obtained by recording from each monitoring device that network connection is established with it, General driving trace is obtained, and then analyze city road network by recording using the vehicle got.

Driving trace acquisition methods provided in an embodiment of the present invention, the executive agent of each step can be server.For example, The server can be a server or the server cluster being made of multiple servers or a cloud meter Calculate service centre.For ease of description, in following each embodiments of the method, only using the executive agent of each step as server into Row is not for example, but form this restriction.

It please refers to Fig.1, it illustrates the flow charts of driving trace acquisition methods provided by one embodiment of the present invention.Such as figure Shown in 1, which can include the following steps.

Step 110, the vehicle of each bayonet is obtained by record, and the vehicle of the bayonet is by recording comprising by the bayonet The vehicles identifications of vehicle, the vehicle pass through the time of the bayonet and the mark of the bayonet.

Optionally, the vehicles identifications of above-mentioned vehicle can be the license plate number of the vehicle.

Step 120, the vehicle comprising effective vehicles identifications is carried out by recording according to vehicle by the time in record Sequence generates the corresponding history driving trace of each effective vehicles identifications according to ranking results.

Wherein, effective vehicles identifications are legal vehicles identifications, and such as effective vehicles identifications are non-deck license plate number.Server is true It is achieved by those of ordinary skill in the art that whether determine vehicles identifications, which be effective vehicles identifications,.It is somebody's turn to do for example, server obtains The corresponding reference picture of vehicles identifications (reference picture can be provided by user) obtains the image of monitoring device shooting at bayonet, The textural characteristics code of this two images is obtained respectively, calculates the similarity between the two textural characteristics codes.If similarity is less than Similarity threshold then judges the vehicles identifications for effective vehicles identifications, and otherwise, which identifies for illegal vehicle.This reality Example is applied to determining whether vehicles identifications are that the modes of effective vehicles identifications no longer repeats one by one.

For each effective vehicles identifications, server obtains all vehicles comprising the vehicles identifications and passes through record.For The vehicle got can first be rejected effective vehicles identifications therein, be passed through in record according still further to vehicle by record, server The sequencing of time is ranked up, and obtains the corresponding history driving trace of the effective vehicles identifications；Alternatively, server first according to Vehicle is sorted, then reject effective vehicles identifications therein by the sequencing of the time in record, obtains effective vehicle mark Know corresponding history driving trace.

Come using effective vehicles identifications as capital * * * * * for example, all vehicles for including capital * * * * * that server is got It is as follows by recording：Bayonet A_2016-11-9 18:09:58_ vehicles capital * * * * *；Bayonet B_2016-11-9 18:11:01_ vehicles Capital * * * * *；Bayonet C_2016-11-9 18:13:13_ vehicles capital * * * * *.Server according to vehicle by being recorded in record when Between sequencing to more than vehicle by record be ranked up, reject sequence after each vehicle pass through effective vehicle in record Mark, the history driving trace for obtaining vehicle capital * * * * * are：A_2016-11-9 18:09:58->B_2016-11-9 18:11: 01->C_2016-11-918:13:13->D_2016-11-11 8:30:01->E_2016-11-11 8:32:01.Wherein, symbol Combination "->" for separate two vehicles pass through record.

Step 130, the stop section in every history driving trace is identified, according to stopping section by the history driving trace Single driving trace is divided into, counts the first frequency of identical single driving trace.

Wherein, the intensive traffic section between adjacent bayonet where when vehicle generation stop over behavior is in above-mentioned stop section, this is adjacent Bayonet is the bayonet that position is adjacent in history driving trace.Single driving trace is that stop over does not occur in driving process of vehicle Track during behavior.

Come with identifying that effective vehicles identifications capital * * * * * correspond to the stop section in history driving trace for example, vehicle The history driving trace of capital * * * * * is：A_2016-11-9 18:09:58->B_2016-11-918:11:01->C_2016-11- 9 18:13:13->D_2016-11-11 8:30:01->E_2016-11-11 8:32:01.Server may recognize that vehicle Stop over behavior has occurred in capital * * * * * between bayonet C and bayonet D, and the section between bayonet C and bayonet D is stop section, then server History driving trace can be divided into two single driving traces, two single driving traces are respectively：1、A->B->C；2、D-> E。

Server counts identical single driving trace after every history driving trace is divided into single driving trace First frequency.For example, single driving trace A->B->C and single driving trace A->B->C->D is differed.For single Driving trace A->B->C counts A- in all single driving traces>B->The number that C occurs obtains A->B->The first frequency of C.

Step 140, general traveling is determined from each single driving trace according to the first frequency of each single driving trace Track.

First frequency is more than the list of frequency threshold value by the first frequency of the identical single driving trace of server statistics, server Secondary driving trace is determined as general driving trace.Wherein, frequency threshold value can be set by system developer, also can be by server root According to the quantity of single driving trace, the quantity set of effective vehicles identifications.The present embodiment does not make the setting of frequency threshold value specific It limits.

Determine frequency threshold value come for example, server can be by single row according to the quantity of single driving trace with server It sails 1/10th of the quantity of track and is determined as frequency threshold value.If server draws the history driving trace of 100 vehicles Get 1000 single driving traces, then frequency threshold value is 100.If single driving trace A->B->The frequency of C is 350, then Server is by single driving trace A->B->C is determined as general driving trace.

There are incidence relations between each bayonet in one general driving trace.For example, when vehicle passes through general driving trace In any bayonet when, the vehicle is higher by the possibility of other bayonets in the general driving trace.For another example, in general traveling In track in the case of the vehicle flowrate sustainable growth of any bayonet, other bayonets increase in the general driving trace possibility Property is higher.For another example, when traffic congestion occurs for any bayonet in general driving trace, other bayonets in the general driving trace The possibility to get congestion is higher.

With general driving trace A->B->C comes for example, the bayonet in general driving trace is bayonet A, bayonet B and card Mouthful C, there are incidence relations between these bayonets.

In conclusion driving trace acquisition methods provided in an embodiment of the present invention, since bayonet divides extensively in city road network Cloth, the vehicle of each bayonet may include that the vehicle by all vehicles by the bayonet by record, can reflect by record Go out the history traveling behavior of vehicle, thus the corresponding history traveling rail of each effective vehicles identifications is generated by recording by vehicle Mark, and after the stop section in identifying every history driving trace, according to the stop section identified by the history row It sails track and is divided into single driving trace, can obtain the single driving trace of most vehicles in city road network；Solves phase City road network analysis result accuracy is analyzed according to the driving trace that global positioning system in each car-mounted terminal records in the technology of pass The problem of low；Achieve the effect that improve city road network precision of analysis.

A is please referred to Fig.2, it illustrates the flow charts for the driving trace acquisition methods that another embodiment of the present invention provides. As shown in Figure 2 A, which can include the following steps.

Step 210, the vehicle of each bayonet is obtained by record, and the vehicle of the bayonet is by recording comprising by the bayonet The vehicles identifications of vehicle, the vehicle pass through the time of the bayonet and the mark of the bayonet.

Step 220, the vehicle comprising effective vehicles identifications is carried out by recording according to vehicle by the time in record Sequence generates the corresponding history driving trace of each effective vehicles identifications according to ranking results.

The explanation of this step 220 can refer to the explanation of step 120.

Step 230, the stop section in every history driving trace is identified, when stopping section as vehicle generation stop over behavior The intensive traffic section between adjacent bayonet.

Optionally, identify that the stop section in a history traveling rail can be realized by several steps as shown in Figure 2 B.

In step 231, the first bayonet adjacent two-by-two in the history driving trace is combined, obtains the combination of the first bayonet Corresponding vehicle is used as the transit time that vehicle combined by first bayonet by the time in record.

For adjacent two-by-two two vehicles of sorting in the history driving trace by record, server will be recorded wherein Two bayonets identify corresponding bayonet and are combined as the first bayonet, and the time wherein recorded is determined as vehicle first is blocked by this The transit time of mouth combination, the vehicle refer to have the vehicle that the history driving trace corresponds to vehicles identifications.

Come by the corresponding history driving traces of capital * * * * * of effective vehicles identifications for example, server obtains the history Two vehicles for sorting adjacent in driving trace can be A_2016-11-9 18 by record by record, this two vehicles: 09:58 and B_2016-11-9 18:11:01.Server is combined bayonet A and B as the first bayonet, by time 2016-11-9 18:09:58 and 2016-11-9 18:11:01 is determined as the transit time that vehicle capital * * * * * pass through bayonet A, B.

In step 232, a period corresponding to above-mentioned transit time is determined, obtain first bayonet combination in the period Longest in corresponding normal pass duration is passed through duration.

The present embodiment does not do specific restriction to the division of period, can be determines according to actual conditions.It optionally, can be by one day It is divided into several periods.For example, 12 duration equal periods will be divided within 24 hours one day, [0:00,2:00) it is first A period, [2:00,4:00) it is second period, other and so on, [22:00,0:00) it is the 12nd period, 2016- [the 0 of 11-9:00,2:00) with [the 0 of 2016-11-10:00,2:00) it is the same period.

Optionally, several periods were divided by one week.For example, it was divided into 84 duration equal periods, star by one week [the 0 of phase one:00,2:00) it is first period, [the 2 of Monday:00,4:00) it is second period, other and so on, star [the 22 of phase day:00,0:00) it is the 84th period.

Server can pass through following several possible embodiments when determining above-mentioned transit time corresponding period It realizes：

In the first possible embodiment, server determines that time sequencing is forward in above-mentioned two transit time The period is determined as an above-mentioned two transit time corresponding period by the period that transit time is located at.

It is respectively 2016-11-9 18 with two transit times:09:58 and 2016-11-9 18:11:01 come for example, By the forward transit time 2016-11-9 18 of time sequencing:09:58 corresponding periods [18:00,20:00) it is determined as the two The transit time corresponding period.

It is when above-mentioned two transit time is located at the same period, the period is true in second of possible embodiment It is set to the above-mentioned two transit time corresponding period；When above-mentioned two transit time is located at two different periods, determine this two Time range between a transit time accounts for the ratio of the two periods, by the ratio high period be determined as above-mentioned two it is current when Between the corresponding period.

Using two transit times as 2016-11-9 19:30:00 and 2016-11-9 22:10:00 come for example, 2016-11-9 19:30:00 corresponding period was [18:00,20:00), 2016-11-9 22:10:00 corresponding period was [22:00,24:00), the time range between two transit times is 2016-11-9 19:30:00~2016-11-922:10: 00.Time range between two transit times accounts for the period [18:00,20:00) ratio is 0.25, accounts for the period [18:00,20: 00) ratio is 0.083, by the period [18:00,20:00) it is determined as the two transit times corresponding period.

Server obtains first bayonet combination in the period after the above-mentioned two transit time corresponding period is determined Longest in corresponding normal pass duration is passed through duration.Wherein, during normal pass a length of driving vehicle do not occur stop, not In the case of driving over the speed limit, the current duration that may be used is combined by the first bayonet, when longest is current during a length of normal pass Maximum value in length.

The combination of each first bayonet can be set in the longest of the day part duration that passes through by system developer, also can be by servicing Device current duration calculation according to used in each vehicle combined in the period by the first bayonet obtains.

In step 233, the current duration which is combined by first bayonet is calculated according to above-mentioned transit time, at this Current duration be more than the longest pass through duration when, the section between first bayonet combination is determined as in the history driving trace Stop section.

Server calculate above-mentioned two transit time difference obtain the vehicle by first bayonet combine it is current when It is long.Using two transit times as 2016-11-9 18:09:58 and 2016-11-9 18:11:01 comes for example, server meter Calculate transit time 2016-11-9 18:09:58 and transit time 2016-11-9 18:11:01 difference obtains the vehicle and passes through Bayonet A, B it is current when a length of 63 seconds.

Server this pass through duration more than the longest pass through duration when, by first bayonet combine between section determine For the stop section in the history driving trace.

For example, using two transit times as 2016-11-9 18:09:58 and 2016-11-9 19:09:58 illustrate Illustrate, a length of 1 hour when vehicle capital * * * * * used in bayonet A, B by passing through, the two transit times corresponding period [18:00,20:00).If bayonet A, B are in the period [18:00,20:00) a length of 80s when longest is current, then server think vehicle Stop over behavior has occurred in capital * * * * * between bayonet A, B, and the section between bayonet A, B is the stop road in the history driving trace Section.

Step 240, which is divided into single traveling by the stop section in the history driving trace Track.

The vehicle of sequence first in the history driving trace is determined as a starting point by server by record, is sorted most Vehicle afterwards is determined as a terminal by record；The preceding vehicle that sorts in the stop section of the history driving trace is passed through Record is determined as the terminal of a upper single driving trace, and the posterior vehicle that sorts is determined as next single by record and travels The starting point of track.For every single driving trace, server is rejected each vehicle in the single driving trace and is passed through in record Time.

For example, the history driving trace for vehicle capital * * * * * as shown in Figure 2 C, the section between bayonet C and bayonet D is Stop section.The history driving trace can be divided into two single driving traces by server, respectively：A_2016-11-9 18:09:58->B_2016-11-9 18:11:01->C_2016-11-9 18:13:13 and D_2016-11-11 8:30:01-> E_2016-11-11 8:32:01.The time that server rejecting wherein records, obtaining two single driving traces is respectively：A-> B->C and D->E.

Step 250, count the first frequency of identical single driving trace, according to the first frequency of each single driving trace from General driving trace is determined in each single driving trace.

The explanation of this step can refer to the explanation of step 140.

Since general driving trace is the rail that is frequently occurred in the single driving trace of most vehicles in city road network Mark, analysis is carried out to city road network for the general driving trace of server by utilizing so that analysis result is more accurate.Server can also base The relationship in city road network between some the intensive traffic sections is analyzed in general driving trace, specifically can this be analyzed based on general driving trace Relational implementation between the bayonet of a little the intensive traffic sections.Server analysis of relationship between bayonet can refer to step 260 to step 280.

Step 260, the second bayonet combination to conform to a predetermined condition in every general driving trace is extracted.

Wherein, predetermined condition includes bayonet in default bayonet item number and/or the combination of the second bayonet in general driving trace In position it is whether continuous.

It is 2 by bayonet item number of predetermined condition, position of the bayonet in general driving trace is continuously come for example, from logical With driving trace A->B->The the second bayonet combination to conform to a predetermined condition is extracted in C：(A, B) and (B, C).

Be again 2 by bayonet item number of predetermined condition, position of the bayonet in general driving trace it is discontinuous, from general traveling Track (A->B->C the second bayonet to conform to a predetermined condition is extracted in) to be combined as (A, C).

It is 2 to come for example, from general driving trace A- by bayonet item number of predetermined condition again>B->It extracts and meets in C The second bayonet combination of predetermined condition：(A, B), (A, C) and (B, C).

Step 270, the support and at least two confidence levels of each second bayonet combination are calculated.

This step can be realized by several steps as shown in Figure 2 D.

Step 271, any second bayonet is combined, counts the second frequency of second bayonet combination.

Identical bayonet, which is identified, comprising bayonet is combined as a kind of second bayonet combination, each second card of server statistics Second frequency of mouth combination.For example, it is a kind of second bayonet that (A, C, B) is combined in the second bayonet combination (A, B, C), which with the second bayonet, Combination, the second bayonet combination (A, B, C) are combined (A, B, D) with the second bayonet and are not combined for same second bayonet.

Step 272, quotient is asked to the quantity of second frequency and general driving trace, using obtained result as second card The support of mouth combination.

It is 3 by bayonet item number of predetermined condition, position of the bayonet in general driving trace is continuously come for example, all General driving trace it is as shown in table 1, the second frequency of the second bayonet combination (A, B, C) is 3, the number of all general driving traces It is 4 to measure, then the support of the second bayonet combination (A, B, C) is 75%.

Table 1

Step 273, it for any bayonet in second bayonet combination, counts and is included in the combination of all second bayonets except should The quantity of other bayonets other than any bayonet asks quotient, using obtained result as this to second frequency and statistical result One bayonet corresponding confidence level in second bayonet combination.

For example, all general driving traces are referring still to table 1, for the card in the second bayonet combination (A, B, C) Mouth A, server need to count the second bayonet combination comprising B, C.Referring still to table 1, server is from the 1st, 2,4 article of general traveling It can extract out the second bayonet combination (A, B, C) and (B, C, D) in track.Therefore, server can be counted comprising bayonet B, C The quantity of second bayonet combination is 6.Since the second frequency of the second bayonet combination (A, B, C) is 3, bayonet A is in the second bayonet group The confidence level for closing (A, B, C) is 50%.

Step 280, support is met into support threshold and at least one confidence level meets the second card of confidence threshold value Mouth combination is determined as the bayonet combination of strong incidence relation.

The support threshold met corresponding to the second bayonet combination of different predetermined conditions is not exactly the same, corresponding confidence It is also not exactly the same to spend threshold value.Generally, the corresponding support threshold of the second bayonet combination for meeting identical predetermined condition and Confidence threshold value is identical.

Wherein, support threshold and confidence threshold value are set by system developer, for example system developer is determining When meeting the corresponding support threshold of the second bayonet combination of same predetermined condition, can it is first empirically determined go out support threshold Value can value range, then verified using a large amount of experimental data, therefrom determine the value of support threshold.

In addition, since traffic of the same a road section in different periods has differences, vehicle do not exceed the speed limit, not stop over behavior In the case of by normal pass duration used in the section there is also difference, by determining that vehicle passes through the first bayonet group The two transit times corresponding period closed obtains first bayonet and combines in the period corresponding normal pass duration most The vehicle is relatively somebody's turn to do by long current duration by the current duration that first bayonet combines compared with the longest passes through duration come determining Whether the section between the combination of the first bayonet is to stop section, improves the accuracy for stopping section identification.

In addition, since the bayonet extracted from the general driving trace of same is there are incidence relation, by extracting every The the second bayonet combination to conform to a predetermined condition in general driving trace, can excavate the different types of bayonet that is mutually related, for example, The binomial that bayonet item number is 2 is associated with bayonet, and bayonet item number is more than 2 bi directional association bayonet；For another example, position in general driving trace It puts and is continuously associated with bayonet, the discontinuous association bayonet in position in general driving trace.

Optionally, obtaining longest of the first bayonet combination in certain period corresponding normal pass duration duration that passes through can lead to The several steps crossed as shown in Figure 3A are realized.

In the step 310, the current duration that each vehicle is combined within the period by first bayonet is obtained.

Two bayonets that server is obtained in the combination of the first bayonet are corresponding when position is adjacent in each history driving trace Two vehicles pass through record.If this two vehicles are corresponding with the period by the time in record, server calculates this two Vehicle obtains a current duration by the difference of time in record.

To obtain vehicle capital * * * * * in the period [18：00,20) it is lifted by the current duration of the first bayonet combination (A, B) Example explanation.The history driving trace of vehicle capital * * * * * is：A_2016-11-918:09:58->B_2016-11-9 18:11:01- >C_2016-11-9 18:13:The position in the history driving trace is adjacent with bayonet B by 13, bayonet A, and server obtains bayonet A Vehicle corresponding with bayonet B passes through record.This two vehicles are respectively 2016-11-9 18 by the transit time in record: 09:58 and 2016-11-918:11:01, with the period [18：00,20) it is corresponding.Server calculates this two vehicles and passes through record The difference of middle time obtains vehicle in the period [18：00,20) pass through the current duration of the first bayonet combination (A, B).

In step 320, the duration section where the current duration of each vehicle is determined, statistics is positioned at each duration section The quantity of current duration, according to statistical result generating probability density function.

For example, duration section can be [0s, 5s), [5s, 10s), [10s, 15s), [15s, 20s) etc., other are successively Analogize.Wherein, the present embodiment does not make the division in duration section specific restriction, when it is implemented, the duration section can be by system Developer sets.

For example, pass through when a length of 11s where duration section for [10s, 15s).

Server determines the duration section where the current duration of each vehicle, and statistics is positioned at the current of each duration section The quantity of duration.During according to statistical result generating probability density function, for each duration section, server will be located at the duration The quantity of the current duration in section and the quotient of the total quantity of current duration are as the corresponding probability in the duration section.

Each duration section of server by utilizing corresponds to probability generating probability density function, and the independent variable of the probability density function is Current duration, dependent variable is probability density.Specifically, server is directed to each duration section, it is corresponding to the duration section general Rate and the length in the duration section ask quotient to obtain the corresponding probability density in the duration section.Server takes according to each duration section Value and the corresponding probability density generating probability density function in each duration section.

In a step 330, hypothetical inspection is carried out to the probability density function, will met in the current duration value of hypothesis Maximum value be determined as the first candidate longest and pass through duration.

The current duration for meeting hypothesis is determined as normal pass duration by server, by the maximum value in normal pass duration It is determined as the first candidate longest to pass through duration.Optionally, it is used aobvious when carrying out hypothetical inspection to probability density function Work property level is usually 0.05 or 0.01, and significance is to be rejected the upper limit value of probability.

Come using significance as 0.05 for example, be certain section as shown in Figure 3B certain period current duration it is general Rate density function, shade therein are region of rejection, and the area of single region of rejection is no more than 0.025.M points in independent variable reference axis Value between Y points meets hypothesis, that is to say, that meets the normal pass duration of hypothesis ranging from：M points value~Y points take Value.Wherein, Y points value is the maximum value for the current duration for meeting hypothesis, and Y point values are determined as the first candidate longest by server Current duration.

In step 340, abnormal current duration is determined using CUSUM algorithms and the probability density function, by the exception The maximum value in current duration except current duration is determined as the second candidate longest and passes through duration.

N point of observation, the value and i-th of the i+1 point of observation of selection are chosen in the independent variable of probability density function The difference of the value of a point of observation is predetermined value δ, the value of n-th of point of observation of selection and above-mentioned each duration section value Maximum value difference be less than δ.Wherein, i, δ and n are positive integer, n>i.

Here, statistic of the present embodiment to be calculated by test statistics for hypothesis testing.In null hypothesis situation Under, this statistic obeys a given probability distribution, and this is quite different in the case where another kind is assumed.If so as to test statistics Value fall except the critical value of above-mentioned distribution, then it is believed that aforementioned null hypothesis may not be correct.For example, in probability density letter Several independents variable passes through and 10 points of observation is chosen in duration, the values of these points of observation may respectively be 5s, 10s, 15s, 20s, 25s, 30s, 35s, 40s, 45s and 50s.For i-th of point of observation, before server calculates, i point of observation is corresponding examines system Metering.When the corresponding test statistics of preceding i point of observation is less than abnormal threshold value, server is by the value of i-th of point of observation It is determined as normal value, otherwise, the value of i-th of point of observation is determined as abnormal value.Wherein, abnormal threshold value is usually by being The developer that unites sets.

In x_iIt is when not being the cut off value in any duration section adjacent two-by-two, the log-likelihood numerical value of i-th of point of observation is true It is set to 0, by the corresponding test statistics T of preceding i-1 point of observation_i-1Value be determined as T_i；In x_iPoint for adjacent duration section During dividing value, the corresponding probability density in posterior duration section that sorts probability density corresponding with preceding duration section of sorting is asked Quotient obtains the likelihood numerical value of i-th of point of observation, and the natural logrithm for calculating the likelihood numerical value obtains log-likelihood numerical value, right using this Number likelihood numerical value test statistics T corresponding with preceding i-1 point of observation_i-1Be added summation, by summed result and numerical value 0 most The corresponding test statistics T of i point of observation before big value is determined as_i.Wherein, sort preceding duration in above-mentioned adjacent duration section The upper limit value in section is equal with the lower limiting value in posterior duration section of sorting, and is the cut off value in the two adjacent duration sections；Work as i When being 1, the corresponding test statistics T of preceding i-1 point of observation₀Value be 0.

The abnormal value determined is determined as the minimum value of the current duration value range of an exception by server, will be determined Go out with upper limit value of the immediate normal value of the exception value as the current duration value range of the exception.That is, [abnormal value, the immediate normal value of the exception value) in current duration be abnormal current duration.

For example, the value of point of observation may respectively be 5s, 10s, 15s, 20s, 25s, 30s, 35s, and others are with such It pushes away.If server determines that the value 25s of the 5th point of observation is normal value, the value 30s of the 6th point of observation takes to be abnormal Value, the value of the 7th, 8 point of observation is also abnormal value, and the value 45s of the 9th point of observation is normal value, then server can Determine [30s, 45s) in current duration be abnormal current duration.

Current duration between numerical value 0 and n-th of point of observation value in addition to abnormal current duration is determined as just by server Maximum value in the normal pass duration is determined as the second candidate longest and passed through duration by normal open row duration.

In step 350, it is the first candidate longest passes through duration and the second candidate longest is passed through in duration maximum value is true It is set to longest of first bayonet combination in the period corresponding normal pass duration to pass through duration.

Fig. 4 A are the block diagrams of driving trace acquisition device provided in one embodiment of the invention.The present embodiment with The driving trace acquisition device in server applied to illustrating.As shown in Figure 4 A, which includes： Acquisition module 410, generation module 420, identification module 430,440 and first determining module 450 of division module.

Acquisition module 410, for obtaining the vehicle of each bayonet by record, the vehicle of the bayonet is included by record to be passed through The vehicles identifications of the vehicle of the bayonet, the vehicle pass through the time of the bayonet and the mark of the bayonet.

Generation module 420, for by the vehicle comprising effective vehicles identifications that acquisition module 410 is got by record by It is ranked up according to vehicle by the time in record, generating the corresponding history of each effective vehicles identifications according to ranking results travels Track.

Identification module 430, for identifying the stop section in every history driving trace of the generation of generation module 420.

The history driving trace is divided into single by division module 440, the stop section identified according to identification module 430 Driving trace counts the first frequency of identical single driving trace.

First determining module 450, it is true from each single driving trace for the first frequency according to each single driving trace Make general driving trace.

Wherein, the intensive traffic section when vehicle generation stop over behavior is in above-mentioned stop section between adjacent bayonet.

In a kind of possible embodiment, as shown in Figure 4 B, Fig. 4 B are the rows provided in another embodiment of the present invention Sail the block diagram of track acquisition device, identification module 430, including：First acquisition unit 431,432 and of second acquisition unit Determination unit 433.

First acquisition unit 431, for being combined for the first bayonet adjacent two-by-two in the history driving trace, obtaining should First bayonet combines corresponding vehicle is used as the transit time that vehicle combined by first bayonet by the time in record.

Second acquisition unit 432, during for determining one corresponding to transit time that first acquisition unit 431 gets Section obtains longest of first bayonet combination in the period corresponding normal pass duration and passes through duration, during the normal pass A length of current duration for not occurring to stop, do not drive over the speed limit.

Determination unit 433, the transit time got according to first acquisition unit 431 calculate the vehicle and first are blocked by this The current duration of mouthful combination, when the longest that the duration that passes through is got more than second acquisition unit 432 passes through duration, by this Section between the combination of one bayonet is determined as the stop section in the history driving trace.

In a kind of possible embodiment, as shown in Figure 4 B, second acquisition unit 432 is used for：Each vehicle is obtained to exist The current duration combined in the period by first bayonet；Determine the duration area where the current duration of above-mentioned each vehicle Between, statistics is located at the quantity of the current duration in each duration section, according to statistical result generating probability density function；It is close to the probability It spends function and carries out hypothetical inspection, the maximum value met in the current duration value of hypothesis is determined as the first candidate longest passes through Duration；Abnormal current duration is determined using CUSUM algorithms and the probability density function, it will be logical except the abnormal current duration Maximum value in row duration is determined as the second candidate longest and passes through duration；By the first candidate longest pass through duration and second it is candidate most Maximum value in long current duration is determined as the longest that first bayonet is combined in the period corresponding normal pass duration and leads to Row duration；Wherein, the independent variable of the probability density function is current duration.

In a kind of possible embodiment, as shown in Figure 4 B, division module 440 is used for：It will be in the history driving trace The vehicle of sequence first is determined as a starting point by record, sorts and is determined as a terminal by record in last vehicle； The preceding vehicle that will sort in the stop section is determined as the terminal of a upper single driving trace by record, and sequence is posterior Vehicle is determined as the starting point of next single driving trace by record.

In a kind of possible embodiment, as shown in Figure 4 B, which further includes：Extraction module 460th, 470 and second determining module 480 of computing module.

Extraction module 460, for extracting the second bayonet to conform to a predetermined condition in every general driving trace combination.

Computing module 470, for calculating the support and at least two of each second bayonet combination of the extraction of extraction module 460 A confidence level.

Second determining module 480, for support being met support threshold and at least one confidence level meets confidence level The second bayonet combination of threshold value is determined as the bayonet combination of strong incidence relation.

Wherein, above-mentioned predetermined condition includes bayonet in default bayonet item number and/or the combination of the second bayonet in the general row Whether the position sailed in track is continuous.

In a kind of possible embodiment, as shown in Figure 4 B, computing module 470, including：Statistic unit 471, first is counted Calculate 472 and second computing unit 473 of unit.

Statistic unit 471 for being combined for any second bayonet, counts the second frequency of second bayonet combination.

First computing unit 472, for the second frequency and the quantity of general driving trace counted on to statistic unit 471 Quotient is asked, the support that obtained result is combined as second bayonet.

Second computing unit 473, for for any bayonet in second bayonet combination, counting all second bayonet groups The quantity of other bayonets in addition to any bayonet is included in conjunction, to the second frequency that statistic unit 471 counts on and statistics As a result ask quotient, using obtained result as any bayonet second bayonet combination in corresponding confidence level.

In conclusion driving trace acquisition device provided in an embodiment of the present invention, since bayonet divides extensively in city road network Cloth, the vehicle of each bayonet may include that the vehicle by all vehicles by the bayonet by record, can reflect by record Go out the history traveling behavior of vehicle, thus the corresponding history traveling rail of each effective vehicles identifications is generated by recording by vehicle Mark, and after the stop section in identifying every history driving trace, according to the stop section identified by the history row It sails track and is divided into single driving trace, can obtain the single driving trace of most vehicles in city road network；Solves phase City road network analysis result accuracy is analyzed according to the driving trace that global positioning system in each car-mounted terminal records in the technology of pass The problem of low；Achieve the effect that improve city road network precision of analysis.

It should be noted that：The driving trace acquisition device provided in above-described embodiment when obtaining driving trace, only with The division progress of above-mentioned each function module, can be as needed and by above-mentioned function distribution by not for example, in practical application Same function module is completed, i.e., the internal structure of server is divided into different function modules, described above complete to complete Portion or partial function.In addition, driving trace acquisition device and driving trace acquisition methods embodiment that above-described embodiment provides Belong to same design, specific implementation process refers to embodiment of the method, and which is not described herein again.

Term " first ", " second " are only used for description purpose, and it is not intended that instruction or hint relative importance or hidden Quantity containing indicated technical characteristic." first " that limits as a result, the feature of " second " can express or implicitly include one A or more this feature.In the description of the present invention, unless otherwise indicated, " multiple " be meant that two or two with On.

The embodiments of the present invention are for illustration only, do not represent the quality of embodiment.

One of ordinary skill in the art will appreciate that hardware can be passed through by realizing all or part of step of above-described embodiment It completes, relevant hardware can also be instructed to complete by program, the program can be stored in a kind of computer-readable In storage medium, storage medium mentioned above can be read-only memory, disk or CD etc..

The foregoing is merely presently preferred embodiments of the present invention, is not intended to limit the invention, it is all the present invention spirit and Within principle, any modification, equivalent replacement, improvement and so on should all be included in the protection scope of the present invention.

Claims

1. a kind of driving trace acquisition methods, which is characterized in that the method includes：

The vehicle of each bayonet is obtained by record, the vehicle of the bayonet includes the vehicle of the vehicle by the bayonet by recording Mark, the vehicle pass through the time of the bayonet and the mark of the bayonet；

Vehicle comprising effective vehicles identifications is ranked up by recording according to vehicle by the time in record, according to sequence As a result the corresponding history driving trace of each effective vehicles identifications is generated；

It identifies the stop section in every history driving trace, is divided into the history driving trace according to the stop section Single driving trace counts the first frequency of identical single driving trace；

General driving trace is determined from each single driving trace according to the first frequency of each single driving trace；

Wherein, the intensive traffic section stopped when section is vehicle generation stop over behavior between adjacent bayonet.

2. the according to the method described in claim 1, it is characterized in that, stop road identified in every history driving trace Section, including：

First bayonet adjacent two-by-two in the history driving trace is combined, first bayonet is obtained and combines corresponding vehicle The transit time that vehicle combined by first bayonet is used as by the time in record；

It determines a period corresponding to the transit time, it is corresponding normal in the period to obtain the first bayonet combination Longest in current duration is passed through duration, a length of current duration for not occurring to stop, do not drive over the speed limit during the normal pass；

According to the current duration that the transit time calculating vehicle is combined by first bayonet, in the current duration More than the longest pass through duration when, by first bayonet combine between section be determined as in the history driving trace Stop section.

3. according to the method described in claim 2, it is characterized in that, described obtain the first bayonet combination in the period pair Longest in the normal pass duration answered is passed through duration, including：

Obtain the current duration that each vehicle is combined within the period by first bayonet；

Determine the duration section where the current duration of each vehicle, statistics is positioned at the number of the current duration in each duration section Amount, according to statistical result generating probability density function；

Hypothetical inspection is carried out to the probability density function, the maximum value met in the current duration value of hypothesis is determined as First candidate longest is passed through duration；

Determine abnormal current duration using accumulation and CUSUM algorithms and the probability density function, by the exception it is current when The maximum value in current duration except length is determined as the second candidate longest and passes through duration；

Described first candidate longest is passed through into duration and the second candidate longest maximum value in duration of passing through is determined as described first Longest of the bayonet combination in the period corresponding normal pass duration is passed through duration；

Wherein, the independent variable of the probability density function is current duration.

4. according to the method described in claim 1, it is characterized in that, described travel rail according to the stop section by the history Mark is divided into single driving trace, including：

The vehicle of sequence first in the history driving trace is determined as a starting point by record, is sorted in last vehicle One terminal is determined as by record；

The preceding vehicle that sorts in the stop section is determined as to the terminal of a upper single driving trace, sequence by record Posterior vehicle is determined as the starting point of next single driving trace by record.

5. according to the method any in Claims 1-4, which is characterized in that in first according to each single driving trace After frequency determines general driving trace from each single driving trace, the method further includes：

Extract the second bayonet combination to conform to a predetermined condition in every general driving trace；

Calculate the support and at least two confidence levels of each second bayonet combination；

Support is met into support threshold and at least one confidence level meets the second bayonet combination of confidence threshold value and is determined as The bayonet combination of strong incidence relation；

Wherein, the predetermined condition includes bayonet in default bayonet item number and/or the combination of the second bayonet in the general traveling Whether the position in track is continuous.

6. the according to the method described in claim 5, it is characterized in that, support and extremely for calculating the combination of each second bayonet Few two confidence levels, including：

Any second bayonet is combined, counts the second frequency of the second bayonet combination；

Quotient is asked to the quantity of second frequency and general driving trace, obtained result is combined as second bayonet Support；

For any bayonet in second bayonet combination, count and included in all second bayonet combinations except any bayonet The quantity of other bayonets in addition asks quotient, using obtained result as any card to second frequency and statistical result Mouth corresponding confidence level in second bayonet combination.

7. a kind of driving trace acquisition device, which is characterized in that described device includes：

Acquisition module, for obtaining the vehicle of each bayonet by record, the vehicle of the bayonet is included by record by described The vehicles identifications of the vehicle of bayonet, the vehicle pass through the time of the bayonet and the mark of the bayonet；

Generation module, for by the vehicle comprising effective vehicles identifications that the acquisition module is got by record according to vehicle It is ranked up by the time in record, the corresponding history driving trace of each effective vehicles identifications is generated according to ranking results；

Identification module, for identifying the stop section in every history driving trace of the generation module generation；

The history driving trace is divided into single by division module, the stop section for being identified according to the identification module Driving trace counts the first frequency of identical single driving trace；

First determining module is determined for the first frequency according to each single driving trace from each single driving trace general Driving trace；

8. device according to claim 7, which is characterized in that the identification module, including：

First acquisition unit, for for adjacent the first bayonet combines two-by-two in the history driving trace, obtaining described the One bayonet combines corresponding vehicle is used as the transit time that vehicle combined by first bayonet by the time in record；

Second acquisition unit for determining a period corresponding to transit time that the first acquisition unit gets, obtains Longest of first bayonet combination in the period corresponding normal pass duration is taken to pass through duration, during the normal pass A length of current duration for not occurring to stop, do not drive over the speed limit；

Determination unit, the transit time got according to the first acquisition unit calculate the vehicle and pass through first bayonet The current duration of combination, when the current duration passes through duration more than the longest that the second acquisition unit is got, by institute State the stop section that the section between the combination of the first bayonet is determined as in the history driving trace.

9. device according to claim 8, which is characterized in that the second acquisition unit is used for：

10. device according to claim 7, which is characterized in that the division module is used for：

11. according to the device any in claim 7 to 10, which is characterized in that described device further includes：

Extraction module, for extracting the second bayonet to conform to a predetermined condition in every general driving trace combination；

Computing module, for calculating the support and at least two confidences of each second bayonet combination of the extraction module extraction Degree；

Second determining module, for support being met support threshold and at least one confidence level meets the of confidence threshold value The combination of two bayonets is determined as the bayonet combination of strong incidence relation；

12. according to the devices described in claim 11, which is characterized in that the computing module, including：

Statistic unit for being combined for any second bayonet, counts the second frequency of the second bayonet combination；

First computing unit, the second frequency and the quantity of general driving trace for being counted on to the statistic unit ask quotient, The support that obtained result is combined as second bayonet；

Second computing unit, for for any bayonet in second bayonet combination, counting in all second bayonet combinations The quantity of other bayonets in addition to any bayonet is included, the second frequency that the statistic unit counts on and statistics are tied Fruit asks quotient, using obtained result as any bayonet second bayonet combination in corresponding confidence level.