CN102607553B - Travel track data-based stroke identification method - Google Patents

Travel track data-based stroke identification method Download PDF

Info

Publication number
CN102607553B
CN102607553B CN201210056545.5A CN201210056545A CN102607553B CN 102607553 B CN102607553 B CN 102607553B CN 201210056545 A CN201210056545 A CN 201210056545A CN 102607553 B CN102607553 B CN 102607553B
Authority
CN
China
Prior art keywords
stop
tracing point
candidate
stop place
place
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Expired - Fee Related
Application number
CN201210056545.5A
Other languages
Chinese (zh)
Other versions
CN102607553A (en
Inventor
张健钦
仇培元
王晏民
徐志洁
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Beijing University of Civil Engineering and Architecture
Original Assignee
Beijing University of Civil Engineering and Architecture
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Beijing University of Civil Engineering and Architecture filed Critical Beijing University of Civil Engineering and Architecture
Priority to CN201210056545.5A priority Critical patent/CN102607553B/en
Publication of CN102607553A publication Critical patent/CN102607553A/en
Application granted granted Critical
Publication of CN102607553B publication Critical patent/CN102607553B/en
Expired - Fee Related legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Abstract

The invention discloses a travel track data-based stroke identification method. The stroke identification method comprises the following steps of: dividing track points through speed, and combining the track points with the speed lower than a certain speed threshold value into candidate remaining positions; and combining the candidate remaining positions by utilizing a distance threshold value and a time threshold value so as to determine a true remaining point. According to the stroke identification method, the problems of positioning drift and jitter of mobile phone positioning track data is solved, and the stroke identification precision is high; when 300 seconds is taken as the time threshold value and 1,100 meters is taken as the distance threshold value, the recall ratio is 87.66 percent, and the precision ratio is 81.56 percent; and meanwhile, GPS (Global Position System) positioning track data can be analyzed through adjusting the distance threshold value and the time threshold value.

Description

A kind of stroke recognition methods based on trip track data
Technical field
The present invention relates to a kind of stroke recognition methods based on trip track data.This method is mainly used in by analyzing mobile phone positioning track data acquisition trip track, but also can be applicable to by analyzing GPS positioning track data acquisition trip track simultaneously.
Background technology
Traditional behavioral activity acquisition of information is investigated as main taking resident trip survey and activity log, carries out by the mode of manual research.The investigation of resident trip survey and activity log has formed a set of complete survey process and specification, at home and abroad adopts for many years, but is also subject to the puzzlement of several problems always, as be subject to anti-person's burden larger, and investigation accuracy is not high, spends huge etc.
Along with the development of wireless communication technique, Internet technology, mobile phone and Portable GPS etc. has the equipment of space orientation ability to be popularized in recent years, can automatically record the space latitude and longitude information in continuous moment after arranging.Data after collection can form complete event trace, greatly enrich and strengthened the reduction effect of trip activity, become another effective way of obtaining resident's crawler behavior data.But meanwhile, the trip data acquisition mode of these passive types often lacks respective attributes information as a reference, cannot directly from tracing point, obtain the trip informations such as end points, travel time, trip mode, trip purpose of going on a journey.Therefore, become the focus of resident trip research field for the data processing of trace information and mining analysis method.
At present, the track collection of urban activity object mainly contains the two kinds of modes in GPS location and cell phone network location.The track data that location technology is obtained just comprises longitude and latitude and the corresponding time information thereof of each tracing point, cannot directly obtain the characteristic information of crawler behavior by data itself, as travel time, trip mode, trip purpose, and deeper mechanics etc.Carry out the basis of above these Information Statistics analytical works, identify exactly two kinds of Activity Types of traveler, be i.e. movable dwell phase and mobile phase.Therefore, the task of trip identification be exactly the track that cannot directly understand be converted into can be cognitive stop place and the movement between each stop place.
As shown in Figure 1, stroke identification is exactly that spatially discrete tracing point is divided into and stops moving point and the large class of mobile moving point two.By stopping position or the position range of moving point when obtaining people and carrying out stop activity, generate people's mobile route by mobile moving point.Like this, the relevant information such as duration, stop activity object that stop can be carried out mining analysis by stop place information, and the information such as trip mode, travel time, trip distance can be extracted from mobile route.
At present, stroke recognizer is started with from identification dwell point, mainly contains heuristic approaches and the large class of clustering procedure two.Heuristic approaches also comprises following several:
(1) based on record gap
The GPS equipment using in early stage vehicle driving investigation does not have battery, need to provide electric power by the vehicle in starting.After vehicle launch, the energising of GPS equipment starts recording track data, and when vehicle stops working, device powers down stops record.Therefore, the track of vehicle of acquisition there will be gap in time, can be used for distinguishing parking behavior.Also can there is of short duration flame-out behavior in vehicle in normal driving process, flame-out the distinguishing of vehicle need to set a time threshold and stop activity time.
(2) based on rest point
The method is mainly started with from speed, investigates the tracing point data of obtaining, and thinks that this tracing point is rest point in the time that the speed of tracing point is 0, and that assembles continuously can judge a stop place.Due to the existence of positioning error and drift, tracing point speed when stop is not always 0, and threshold speed need to be set, conventionally taking a speed that is less than walking as standard.
(3) based on missing point
In the situation that underground or buildings block, GPS equipment cannot receive gps signal and produce shortage of data, the time interval interval greater than setting between two adjacent track points that collect.For this situation, by calculating the speed between these two points, and with the velocity ratio of several tracing points of 2 front and back, while judging tracing point disappearance, there is stop activity still mobile movable.
(4) based on direction character
Can there is the change of direction in some stop activity in short-term, pick people, fetch and deliver goods etc. as stopped in the time starting or finish, and can change judgement by the direction of several tracing points of identification and stop.Du is divided into stop the doubtful stop of determining for a long time stop and short time, and whether the direction of investigating doubtful stop changes further checking and stop.
(5) based on road network
Calculate the distance of tracing point and road network, obtain the tracing point that departs from road network, in conjunction with stopping duration, these points are done to further judgement.
Track data generally gathers in the mode of constant duration, will have a large amount of tracing points and be gathered near a certain position, therefore can use the method for cluster to identify while therefore there is stop activity.Clustering procedure also comprises following methods:
(1) based on K-mean cluster
The method first will be determined two parameters: form the minimum track of bunch count n and cluster radius d.From first tracing point, calculate the ultimate range between any two points in n tracing point, if be less than d, these tracing points form one bunch, i.e. a stop place.Afterwards, calculate the distance of next tracing point and this bunch of central point, if be less than d/2, tracing point adds this bunch, otherwise the cluster process of this bunch finishes.Thereby repeat above process, until all tracing points are finished dealing with, cluster goes out several stop places.
(2) based on DBSCAN cluster
Similar to K-mean cluster, also need to determine track count n and cluster radius d.Calculate the tracing point quantity within the scope of each some d, if be less than n, think that this point is noise, otherwise these tracing points form one bunch.If there is coincidence between bunch, merge intersect bunch, finally form several stop places.Record when the method hypothesis tracing point waits all the time, is subject to the impact of shortage of data.
Heuristic approaches is based upon researcher on the experiential basis of the understanding of trip track data and individual's trip rule, sets and continue to optimize multiple rule and parameters for identifying, and reaches the object of stroke identification.The method is pressed close to the experience impression of real world, and required rule and parameter are visual and clear, and reasonably setting can obtain good recognition result.But heuristic approaches is more intense to the specific aim of data, if the obtain manner of track data and feature change, existing method will be no longer applicable.Clustering algorithm to the dependence of existing knowledge experience and the feature of data own a little less than, there is good adaptability, but no matter be that K-average or DBSCAN cluster are poor to the treatment effect of noise and long distance drift, easily will once stop and be divided into repeatedly stop, accuracy of identification is not high.
The trip track data that above-mentioned heuristic approaches is obtained mainly for GPS locator meams, is difficult to be applied to the stroke discriminance analysis of mobile phone positioning track data.GPS location and mobile phone are located the track data that these two kinds of obtain manners obtain and are had a great difference aspect track feature location precision, GPS positioning track data positioning precision is higher, be not prone to the drift of long distance, and mobile phone positioning track data positioning precision and drift features all depend on the distribution density of mobile base station, larger away from the distance of intown region drift, sometimes even may there is location drift and the shake of one, two kilometer; Most people's mobile phone does not generally shut down by day or somebody's whole day is not shut down, and the method based on record gap is greatly limited; Mobile phone signal lid scope in city is more comprehensive, is also not easy to occur the situation of track disappearance in buildings, judges that by missing point the method availability stopping declines.Therefore, also inapplicable to mobile phone positioning track for the recognition methods of GPS track.
Summary of the invention
The present invention has designed and developed a kind of stroke recognition methods based on trip track data.Negotiation speed of the present invention is divided tracing point, and speed is merged into candidate stop place lower than the tracing point below certain speed threshold value, and recycling distance and time threshold merge candidate stop place, thereby determine real dwell point.Said method has solved the location drift of mobile phone positioning track data and the problem of shake, and stroke accuracy of identification is high; Meanwhile, the method can also realize the analysis to GPS positioning track data by adjusting distance and time threshold.
Technical scheme provided by the invention is:
Based on a stroke recognition methods for trip track data, comprise the following steps:
The speed of step 1, calculating tracing point;
Step 2, by multiple adjacent speed all the tracing point below threshold speed merge into a candidate stop place, wherein, the stop duration of described candidate stop place is that in described multiple tracing point, first tracing point arrives the time interval between last tracing point;
Step 3, in the time that in the center of multiple candidates stop place and described multiple candidates stop place, the distance between any candidate stop place is less than distance threshold, and, when between zero hour of the stop duration of first candidate stop place in described multiple candidates stop place to the finish time of the stop duration of last candidate stop place interval greater than time threshold time, the center of described multiple candidates stop place is judged to be to dwell point;
The time interval in step 4, described multiple candidates stop place between the zero hour to the finish time of the stop duration of last candidate stop place of the stop duration of first candidate stop place is the stop duration at described dwell point.
Preferably, in the described stroke recognition methods based on trip track data, described step 3 is achieved in the following ways,
(1) using first candidate stop place in all candidates to be determined stop place as stopping sequence, wherein, described first candidate stop place is the center that stops sequence,
(2) in the time being positioned at distance that first candidate stop place at rear of described stop sequence arrives the center of described stop sequence and being less than distance threshold, described stop sequence is put into in first candidate stop place at the described rear that is positioned at described stop sequence, redefine the center of described stop sequence
(3) repeat (2), until when being positioned at distance that first candidate stop place at described stop sequence rear arrives the center of described stop sequence and being greater than distance threshold, when between zero hour of the stop duration of first candidate stop place in described stop sequence to the finish time of the stop duration of last candidate stop place interval greater than time threshold time, the center that stops sequence described in (2) is dwell point.
Preferably, in the described stroke recognition methods based on trip track data, in described step 2, the described multiple adjacent speed all tracing point below threshold speed is merged into a candidate stop place, be achieved in the following ways,
(1) calculate successively the average coordinates (X of two adjacent track points in described candidate stop place (i, i+1), y (i, i+1)),
(2) calculate successively the time interval Δ t between described two adjacent track points (i, i+1), and the stop duration Stay ' of described candidate stop place. the ratio wight between Δ t (i, i+1),
(3) calculate the coordinate (Stay ' .x, Stay ' .y) of described candidate stop place: Stay ′ . x = 1 n - 1 Σ 1 n - 1 wight ( i , i + 1 ) · x ( i , i + 1 ) , Stay ′ . y = 1 n - 1 Σ 1 n - 1 wight ( i , i + 1 ) · y ( i , i + 1 ) .
Preferably, in the described stroke recognition methods based on trip track data, in described step 3, the center that stops sequence described in (2) is achieved in the following ways,
(1) calculating is positioned at the stop duration Stay ' of first candidate stop place at the rear of described stop sequence i. the ratio wight between the time interval Sq. Δ t in Δ t and described stop sequence between the zero hour of the stop duration of first candidate stop place to the finish time of the stop duration of last candidate stop place i,
(2) calculate the centre coordinate (Sq.x, Sq.y) of described stop sequence:
Sq.x＝wight i·Stay′ i.x+(1-wight i)·Sq.x，
Sq.y＝wight i·Stay′ i.y+(1-wight i)·Sq.y。
Preferably, in the described stroke recognition methods based on trip track data, also include
Step 5, in described step 3, (3) when being positioned at distance that first candidate stop place at described stop sequence rear arrives the center of described stop sequence in and being greater than distance threshold, in the time that the time interval between the zero hour to the finish time of the stop duration of last candidate stop place of the stop duration of first candidate stop place in described stop sequence is less than time threshold, the all tracing points that stop described in (2) between zero hour to the finish time of the stop duration of last candidate stop place of stop duration of first candidate stop place in sequence are all judged to be transfer point, and in described stop sequence, all tracing points between finish time to the zero hour of the stop duration of first candidate stop place at described stop sequence rear of the stop duration of last candidate stop place are all judged to be transfer point.
Preferably, in the described stroke recognition methods based on trip track data, in described step 4, tracing points all between the zero hour of the stop duration of first candidate stop place in described multiple candidates stop place to the finish time of the stop duration of last candidate stop place is deleted.
Preferably, in the described stroke recognition methods based on trip track data, in described step 1, calculate the average velocity of tracing point.
Preferably, in the described stroke recognition methods based on trip track data, in described step 1, calculate the average velocity of tracing point, be achieved in the following ways,
At least one tracing point is chosen respectively at front and rear at current tracing point, calculate from first tracing point to the air line distance last tracing point, calculate first tracing point to the time interval between last tracing point, the average velocity of current tracing point is by obtaining to the time interval last tracing point from described first tracing point to the air line distance last tracing point divided by described first tracing point.
Preferably, in the described stroke recognition methods based on trip track data, in described step 1, calculate the average velocity of tracing point, be achieved in the following ways,
At least one tracing point is chosen respectively at front and rear at current tracing point, calculate respectively two air line distances between adjacent track point, and calculate the air line distance sum between all two adjacent track points in selected tracing point, calculate first tracing point to the time interval between last tracing point, the average velocity of current tracing point obtains to the time interval between last tracing point divided by described first tracing point by the air line distance sum between all two adjacent track points in described selected tracing point.
Preferably, in the described stroke recognition methods based on trip track data, described time threshold is 300 seconds, and described distance threshold is 1100 meters.
Stroke recognition methods based on trip track data of the present invention, Negotiation speed is divided tracing point, and speed is merged into candidate stop place lower than the tracing point below certain speed threshold value, recycling distance threshold and time threshold merge candidate stop place, thereby determine real dwell point.Said method has solved the location drift of mobile phone positioning track data and the problem of shake, and stroke accuracy of identification is high.When time threshold is got 300 seconds, when distance threshold is got 1100 meters, recall ratio is 87.66%, and precision ratio is 81.56%.Meanwhile, the method can also realize the analysis to GPS positioning track data by adjusting distance and time threshold.
Brief description of the drawings
Fig. 1 is stroke identification schematic diagram;
Fig. 2 is that the speed of tracing point is calculated schematic diagram;
Fig. 3 is the schematic diagram that in tracing point, candidate's dwell point is merged into candidate stop place;
Fig. 4 is dwell point recognizer process flow diagram
Fig. 5 is the D prism map of the recall ratio based on different distance threshold value and time threshold in the stroke recognition methods based on trip track data of the present invention;
Fig. 6 is the D prism map of the precision ratio based on different distance threshold value and time threshold in the stroke recognition methods based on trip track data of the present invention;
Fig. 7 is the visual schematic diagram of individual traveler stroke recognition result, wherein, and the space-time path of Fig. 7 (a) for being obtained by initial trace data.(b) be the space-time path of drawing via after stroke identification, the straight line portion identifying through bracket represents the dwell phase identifying, and the bending part between two sections of straight line portioies represents the mobile phase identifying.
Fig. 8 is the conceptual model hierarchical structure based on space-time track data;
Fig. 9 is the relationship map figure in the logical model based on space-time track data;
Figure 10 is the Visualization figure of single traveler track in resident trip track visual analyzing mining prototype system;
Figure 11 is the Visualization figure of multiple traveler tracks in resident trip track visual analyzing mining prototype system;
Figure 12 is the Visualization of the multiple date tracks of single traveler in resident trip track visual analyzing mining prototype system;
Figure 13 be in resident trip track visual analyzing mining prototype system multiple travelers in of even date track distribution situation.
Embodiment
Below in conjunction with accompanying drawing, the present invention is described in further detail, to make those skilled in the art can implement according to this with reference to instructions word.
The invention provides a kind of stroke recognition methods based on trip track data, comprise the following steps:
The speed of step 1, calculating tracing point;
Step 2, by multiple adjacent speed all the tracing point below threshold speed merge into a candidate stop place, wherein, the stop duration of described candidate stop place is that in described multiple tracing point, first tracing point arrives the time interval between last tracing point;
Step 3, in the time that in the center of multiple candidates stop place and described multiple candidates stop place, the distance between any candidate stop place is less than distance threshold, and, when between zero hour of the stop duration of first candidate stop place in described multiple candidates stop place to the finish time of the stop duration of last candidate stop place interval greater than time threshold time, the center of described multiple candidates stop place is judged to be to dwell point;
The time interval in step 4, described multiple candidates stop place between the zero hour to the finish time of the stop duration of last candidate stop place of the stop duration of first candidate stop place is the stop duration at described dwell point.
In the described stroke recognition methods based on trip track data, described step 3 is achieved in the following ways,
(1) using first candidate stop place in all candidates to be determined stop place as stopping sequence, wherein, described first candidate stop place is the center that stops sequence,
(2) in the time being positioned at distance that first candidate stop place at rear of described stop sequence arrives the center of described stop sequence and being less than distance threshold, described stop sequence is put into in first candidate stop place at the described rear that is positioned at described stop sequence, redefine the center of described stop sequence
(3) repeat (2), until when being positioned at distance that first candidate stop place at described stop sequence rear arrives the center of described stop sequence and being greater than distance threshold, when between zero hour of the stop duration of first candidate stop place in described stop sequence to the finish time of the stop duration of last candidate stop place interval greater than time threshold time, the center that stops sequence described in (2) is dwell point.
In the described stroke recognition methods based on trip track data, in described step 2, the described multiple adjacent speed all tracing point below threshold speed is merged into a candidate stop place, be achieved in the following ways,
(1) calculate successively the average coordinates (X of two adjacent track points in described candidate stop place (i, i+1), y (i, i+1)),
(2) calculate successively the time interval Δ t between described two adjacent track points (i, i+1), and the stop duration Stay ' of described candidate stop place. the ratio wight between Δ t (i, i+1),
(3) calculate the coordinate (Stay ' .x, Stay ' .y) of described candidate stop place: Stay ′ . x = 1 n - 1 Σ 1 n - 1 wight ( i , i + 1 ) · x ( i , i + 1 ) , Stay ′ . y = 1 n - 1 Σ 1 n - 1 wight ( i , i + 1 ) · y ( i , i + 1 ) .
In the described stroke recognition methods based on trip track data, in described step 3, the center that stops sequence described in (2) is achieved in the following ways,
(1) calculating is positioned at the stop duration Stay ' of first candidate stop place at the rear of described stop sequence i. the ratio wight between the time interval Sq. Δ t in Δ t and described stop sequence between the zero hour of the stop duration of first candidate stop place to the finish time of the stop duration of last candidate stop place i,
(2) calculate the centre coordinate (Sq.x, Sq.y) of described stop sequence:
Sq.x＝wight i·Stay′ i.x+(1-wight i)·Sq.x，
Sq.y＝wight i·Stay′ i.y+(1-wight i)·Sq.y。
In the described stroke recognition methods based on trip track data, also include step 5, in described step 3, (3) when being positioned at distance that first candidate stop place at described stop sequence rear arrives the center of described stop sequence in and being greater than distance threshold, in the time that the time interval between the zero hour to the finish time of the stop duration of last candidate stop place of the stop duration of first candidate stop place in described stop sequence is less than time threshold, the all tracing points that stop described in (2) between zero hour to the finish time of the stop duration of last candidate stop place of stop duration of first candidate stop place in sequence are all judged to be transfer point, and in described stop sequence, all tracing points between finish time to the zero hour of the stop duration of first candidate stop place at described stop sequence rear of the stop duration of last candidate stop place are all judged to be transfer point.
In the described stroke recognition methods based on trip track data, in described step 4, tracing points all between the zero hour of the stop duration of first candidate stop place in described multiple candidates stop place to the finish time of the stop duration of last candidate stop place is deleted.
In the described stroke recognition methods based on trip track data, in described step 1, calculate the average velocity of tracing point.
In the described stroke recognition methods based on trip track data, in described step 1, calculate the average velocity of tracing point, be achieved in the following ways, at least one tracing point is chosen respectively at front and rear at current tracing point, calculate from first tracing point to the air line distance last tracing point, calculate first tracing point to the time interval between last tracing point, the average velocity of current tracing point is by obtaining to the time interval last tracing point from described first tracing point to the air line distance last tracing point divided by described first tracing point.
In the described stroke recognition methods based on trip track data, in described step 1, calculate the average velocity of tracing point, be achieved in the following ways, at least one tracing point is chosen respectively at front and rear at current tracing point, calculate respectively two air line distances between adjacent track point, and calculate the air line distance sum between all two adjacent track points in selected tracing point, calculate first tracing point to the time interval between last tracing point, the average velocity of current tracing point obtains to the time interval between last tracing point divided by described first tracing point by the air line distance sum between all two adjacent track points in described selected tracing point.
In the described stroke recognition methods based on trip track data, described time threshold is 300 seconds, and described distance threshold is 1100 meters.
Method of the present invention can be divided into three parts: (1) speed is calculated; (2) candidate stop place generates; (3) dwell point identification.
One, speed is calculated
In the initial trace data that obtain, do not comprise velocity information, the first step of algorithm need to calculate the speed of passerby at each tracing point according to longitude, latitude and the temporal information of tracing point record.The instantaneous velocity of stricti jurise calculates more difficult and complicated, therefore considers to replace with the average velocity on one section of track at tracing point place.
For GPS positioning track data, its positioning precision is higher, is not prone to the drift of long distance, and the average velocity on the path that the speed of tracing point is made up of this tracing point and former and later two tracing points that are attached thereto replaces, as the tracing point p in Fig. 2 3, its speed computing formula is as follows:
p 3 . v = D ( 2,3 ) + D ( 3,4 ) Δt ( 2,3 ) + Δt ( 3,4 )
In formula, p 3.v---tracing point p 3speed;
D (i, j)---tracing point p iwith tracing point p jbetween distance;
Δ t (2,3)---tracing point p iwith tracing point p jbetween the time interval.
For mobile phone positioning track data, its positioning precision and drift features depend on the distribution density of mobile base station, larger away from the distance of intown region drift, sometimes may occur location drift and the shake of one, two kilometer.For above problem, attempt adopting the air line distance between tracing point to replace trajectory path to calculate apart from participation speed herein.As shown in Figure 2, calculating tracing point p 3speed time, no longer calculate tracing point p 1, p 2, p 3, p 4, p 5between neighbor distance sum, directly calculate tracing point p 1, p 5between air line distance as traveler at t 1, t 5the path of process between the moment.Computing formula is as follows:
p 3 . v = D ( 1,5 ) Δt ( 1,2 ) + Δt ( 2,3 ) + Δt ( 3,4 ) + Δt ( 4,5 )
By result of calculation and the contrast of trip actual conditions, the velocity characteristic that uses the method to calculate is more consistent with the translational speed that actual trip activity occurs, and can well improve the impact of alignment jitter on speed result of calculation.
Two, candidate stop place generates
According to the speed calculating, tracing point is divided into candidate's dwell point and candidate's transfer point two classes, and continuous candidate's dwell point is merged into candidate stop place, to carry out next step stop judgement.Specifically comprise the work of two aspects: the calculating of the merging of candidate's dwell point and candidate stop place coordinate.
(1) candidate's dwell point merges
The division of candidate's dwell point and candidate's transfer point mainly relies on the threshold speed of setting, and threshold speed is generally got the minimum speed lower limit in resident trip mode, the i.e. lower velocity limit of manner of walking.Normal person's walking speed generally, between thousand ms/h of 3-6, that is to say that the jogging speed of manner of walking is about 0.8m/s.The impact of the consideration speed error of calculation, and on the basis of pre-stage test, get 1 meter per second herein and as threshold speed, tracing point is classified, candidate's dwell point and candidate's transfer point two classes be divided into.Afterwards, two above continuous candidate's dwell point ps are merged into candidate stop place Stay ', candidate stop place is the position that stop activity likely occurs traveler, and whether being defined as dwell point needs follow-up work to do further processing.This process is illustrated as Fig. 3.
(2) coordinate of candidate stop place calculates
Candidate's dwell point of merging into candidate stop place spatially can not be completely overlapping, need to calculate the stop center that can represent each candidate stop place, the namely coordinate of candidate stop place according to the coordinate of these candidate's dwell points.Under the prerequisite of track data constant duration record, can obtain by calculating the average coordinates of each candidate's dwell point the coordinate at the center that stops, but the situation of shortage of data can occur in practice, result of calculation is impacted.Therefore, adopt the coordinate of time-weighted mode calculated candidate stop place herein.
First calculate, successively continuous two candidate's dwell point ps i, ps i+1average coordinates (X (i, i+1), y (i, i+1)):
x ( i , i + 1 ) = ps i . x + ps i + 1 . x 2
y ( i , i + 1 ) = ps i . y + ps i + 1 . y 2
In formula, ps i.x---candidate's dwell point ps ilongitude coordinate;
Ps i.y---candidate's dwell point ps ilatitude coordinate;
Afterwards, by the time interval Δ t between two candidate's dwell points (i, i+1)stop duration Stay ' with whole candidate stop place. the ratio of Δ t is as the weight wight of this average coordinates (i, i+1):
wight ( i , i + 1 ) = Δt ( i , i + 1 ) Stay ′ . Δt
Finally, the coordinate of calculated candidate stop place:
Stay ′ . x = 1 n - 1 Σ 1 n - 1 wight ( i , i + 1 ) · x ( i , i + 1 )
Stay ′ . y = 1 n - 1 Σ 1 n - 1 wight ( i , i + 1 ) · y ( i , i + 1 )
In formula, Stay ' .x---the longitude coordinate of candidate stop place Stay ';
Stay ' .y---the latitude coordinate of candidate stop place Stay '.
Three, dwell point identification
Candidate's dwell point is merged into behind candidate stop place, need to be further analyzed, just can obtain the dwell point of traveler generation activity.Mainly by investigate apart from discrete time two factors candidate stop place is accepted or rejected and is merged, algorithm flow chart is shown in Fig. 4, algorithm steps specifically describe as follows:
Step 1, read first candidate stop place Stay ' 1, put it into and stop sequence Sq, by Stay ' 1coordinate as the centre coordinate that stops sequence Sq.
Step 2, judge whether the candidate stop place of not reading in addition, if so, to read next candidate stop place Stay ' i, calculate Stay ' icentre coordinate and the distance B of Sq centre coordinate (i, Sq), forward step 3 to; If not, forward step 4 to.
Step 3, D (i, Sq)whether be less than the distance threshold Td of setting, if so, by Stay ' iput into and stop sequence Sq, recalculate the centre coordinate that stops sequence Sq, forward step 2 to; If not, forward step 4 to.
Step 4, calculate the stop Sq.st zero hour of Sq and stop the time interval Sq. Δ t of the Sq.et finish time, time interval Sq. Δ t is herein the time interval stopping between zero hour to the finish time of the stop duration of last candidate stop place of the stop duration of first candidate stop place in sequence.
Whether step 5, Sq. Δ t are greater than the time threshold Tt of setting, if, dwell point is merged in candidate stop place in Sq: (1) by be included in tracing point between moment Sq.st and the moment Sq.et of Sq delete (2) by the coordinate of tracing points all between moment Sq.st and moment Sq.et all with the centre coordinate replacement of Sq, forward step 6 to; If not, the candidate stop place in Sq does not form dwell point, and between moment Sq.st and moment Sq.et, all tracing points are all judged to be transfer point, judge in Sq whether comprise last candidate stop place, if so, finish this dwell point judgement, if not, forward step 6 to.
-step 6. empties the candidate stop place in Sq, by Stay ' iput into Sq, by Stay ' istop centre coordinate as the centre coordinate that stops sequence Sq, forward step 2 to.When not comprising last candidate stop place in Sq, also to continue the judgement to candidate stop place, to identify new dwell point, so at Stay ' i-1with Stay ' ibetween candidate's transfer point be also judged as transfer point.
While recalculating the centre coordinate that stops sequence Sq in above-mentioned steps 2, adopt time-weighted method, computing formula is as follows:
wight i = Stay i ′ · Δt Sq . et - Sq . st
Sq.x＝wight i·Stay′ i.x+(1-wight i)·Sq.x
Sq.y＝wight i·Stay′ i.y+(1-wight i)·Sq.y
In formula, wight i---candidate stop place Stay ' iweighted value;
Sq.st---stop the stop start time (zero hour of the stop duration of first candidate stop place) of sequence Sq;
Sq.et---stop the stop end time (finish time of the stop duration of last candidate stop place) of sequence Sq.
Sq.x---stop sequence Sq central point longitude coordinate;
Sq.y---stop sequence Sq central point latitude coordinate.
Four, evaluation index
The index of evaluating stroke recognizer recognition effect mainly comprises recall ratio and precision ratio, and two concept definitions are as follows:
Recall ratio R = DS S
Precision ratio P = DS D
Wherein, DS is the true dwell point number identifying, and S is actual true dwell point number, and D is the dwell point number identifying.
Five, recognition result
Collect 8 volunteers the mobile phone positioning track data of totally 20 days as the research object of mobile phone positioning track data stroke identification.Obtain asking volunteer to recall stroke on the same day after data, determine true dwell point number thereby fill in activity log in conjunction with trip track.
The experience that first design of time threshold and distance threshold obtains with reference to existing research, taking 120 seconds and 200 meters as initial value, then increases taking 60 seconds and 100 meters as unit successively, and maximum is increased to 600 seconds and 1500 meters.Like this, in experiment, have 9 time thresholds and 14 distance thresholds, can obtain 126 groups of threshold value combinations, the dwell point that each combination is identified and actual dwell point comparison, calculate respectively recall ratio and the precision ratio of each combination, obtain recall ratio and precision ratio D prism map as shown in Figure 5 and Figure 6.
As can be known from Fig. 5 and Fig. 6, along with the increase of time threshold and distance threshold, the recall ratio of dwell point identification declines and precision ratio rising, main cause is that less threshold value can be isolated as repeatedly once stopping, larger threshold value can by time or space, occur near stop merge into once stop.By comparative analysis, time threshold is got 300 seconds, when distance threshold is got 1100 meters, can obtain better effects to the stroke identification of mobile phone positioning track data, and recall ratio is 87.66%, and precision ratio is 81.56%.
For GPS positioning track data, time threshold can be chosen 300 seconds, and distance threshold can be chosen 200 meters, and the effect of stroke identification is better, and recall ratio is 88.14%, and precision ratio is 83.25%.
Fig. 7 is the visual signal of individual traveler stroke recognition result.The space-time path of Fig. 7 (a) for being obtained by initial trace data, Fig. 7 (b) is the space-time path of drawing via after stroke identification, straight line portion represents the dwell phase identifying, and the bending part between two sections of straight line portioies represents the mobile phase identifying.
Method of the present invention has good result to overcoming mobile phone locator meams in the static drift on a large scale producing when movable.
Mobile phone positioning track data obtain by installation site logging software on the volunteer's mobile phone recruiting.The software of installing is the Mobiletrack software of ant foot net (http://mymobiletrack.com/) exploitation, and the intelligent mobile phone systems such as this software support WinCE, Symbian S60 and Android adopt Cell-ID localization method.Collect altogether 8 volunteers mobile phone positioning track data of totally 20 days, acquisition time interval is similarly 1 minute, obtains the text data of kml form after base station information processing is carried out in its website.Mainly comprise time, longitude and the latitude information of position record, also need to add the information such as recording key, volunteer's numbering so that contraposition put is distinguished.
The track of traveler is a kind of " space-time track ", storage for this track data will take into full account its space-time characterisation, ensure the relevance between spatial information, temporal information and attribute information, best mode is to design and set up the storage of a Spatio-Temporal Data Model for Spatial for track.
Each element of the mathematical form of expressing for track is represented to (as Fig. 8) with a hierarchical chart, thereby construct a conceptual model.
The object of design logic model is to occur from conceptual model conversion the database logic structure that class relationship type database can be processed, and makes these structures can meet the requirement of user at aspects such as function, performance, integrality, consistance and extensibilities.Fig. 9 is by the relationship map figure of the logical model obtaining after conceptual model conversion.
After conversion, in model, mainly contain four kinds of entity objects: traveler, trip, motion track point and stop.A traveler and basic personal information thereof in the investigation of traveler (PERSON) object encoding, the attribute comprising has: traveler coding (PID), traveler name (PName), sex (Sex), income (Salary), home address (HAddress), driving license situation (PLic), place of working (WAddress) and occupation (Occupation).Wherein PID is primary key, in order to identify different travelers.Certain trip activity of trip (TRIP) object encoding traveler and the association attributes of this trip, mainly contain: trip coding (TID), trip date (Date), traveler key word (TPID), trip sequence number (TSequence), trip purpose (TType) and trip mode (TMode).Wherein TID is primary key, in order to identify trip activity each time.Certain tracing point and time-space attribute information thereof that tracing point (TRACK_POINT) object encoding trip activity is experienced, mainly contain: tracing point coding (TpID), trip key word (TpTID), tracing point moment (TpTime), tracing point longitude coordinate (TpLon) and latitude coordinate (TpLat).Wherein TpID is primary key, in order to identify different tracing points.Stop important playground and space-time and the attribute information of (STAY) object encoding, mainly contain: stop coding (SID), traveler key word (SPID), stop start time (SSTime), stop end time (SETime), dwell point longitude coordinate (SLon), dwell point latitude coordinate (SLat), stop Activity Type (SType).Wherein LID is primary key, in order to indicate different places.
After conceptual model is changed, need to be according to the each tables of data of database design of selecting.Select the database of Oracle as track storage.According to the logical model of design and the traveler track data of acquisition, for track data has designed 4 kinds of tables of data: tracing point tables of data, Traveler Information table, the stop hotlist of each day and the trip hotlist of each day of each day.
(1) tracing point tables of data.This table is mainly stored tracing point positional informations whole in a day, comprise location point numbering, traveler numbering, moment, longitude, latitude, with the information such as a upper location point distance, speed, direction and trip activity SN.The field design of table is as shown in table 1.
Table 1 tracing point tables of data
(2) Traveler Information table.This table is mainly stored the relevant information of traveler, mainly contains traveler numbering, traveler name, sex, income, residence information, residence coordinate, driving license situation, occupation, place of working and place of working coordinate etc.The field design of table is as shown in table 2.
Table 2 Traveler Information table
(3) stop hotlist.This table is mainly stored stop activity for information about, mainly contains traveler numbering, stops the zero hour, stops the finish time, stops location point numbering corresponding to the zero hour, stops Activity Type etc.The field design of table is as shown in table 3.
Table 3 stops hotlist
(4) trip hotlist.This table is mainly stored trip activity for information about, mainly contains traveler numbering, trip activity SN, trip purpose and trip mode etc.The field design of table is as shown in table 4.
Table 4 hotlist of going on a journey
By above-mentioned design, set up a database for space-time track data.Further by using C# programming language and realizing resident trip track visual analyzing mining prototype system in conjunction with the secondary development based on ArcGIS Engine.Resident trip track visual analyzing mining prototype system is realized in multiple situation visual to resident trip track, sees Figure 10-13.
In the axSceneControl scene of ArcGIS Engine, three-dimensional main longitude, latitude and the height (x representing in geographical space, y, z), height z is wherein used for representing moment t, and the three-dimensional scenic of geographical space just can be converted into time space three-dimensional scenic so, represents respectively longitude, latitude and moment (x, y, t).
In the time of specific implementation, create ArcGIS three-dimensional type point according to the longitude of the tracing point obtaining, latitude and moment record from database, and the AddElement method providing by axSceneControl scene control is added in axSceneControl scene, realize the demonstration of tracing point.But the track of traveler is continuous on time and space, no matter be that each tracing point is connected into space-time path, still again mark space-time path by the mode of dynamic segmentation, all need known trajectory point interpolation to obtain the coordinate of other tracing points.For example, obtain the spacetime coordinates (x of unknown tracing point i, y i, t i), can be by known two tracing point (x m, y m, t m), (x n, y n, t n), t m< t i< t nspacetime coordinates calculate, generally adopt linear interpolation formula:
x i = ( t i - t m t n - t m ) ( x n - x m ) + x m
y i = ( t i - t m t n - t m ) ( y n - y m ) + y m
First prototype system provides four basic functions, comprises that the loading of the management of track data and inquiry, track and demonstration, planar movement and track time slice show.Data mining, the result displaying etc. carried out afterwards can be expanded realization on these basic functions.
The management of track data and query function are to provide the basic operations such as importing, maintenance and the deletion of track urtext data in the mode of graphical interfaces.As the core to database manipulation, also encapsulate the SQL statement for track data inquiry simultaneously, simplified querying command, be convenient to calling and developing of other functions.
Track loads and shows that this function has realized the track visual presentation in three-dimensional scenic.Select to show traveler numbering and the date of track by dialog box.Traveler numbering both can have been selected by numbered list (single choice or multiselect), also can the interested traveler numbering of manual data.Equally, the date both can specify a certain exact date, also can specify and select to show the dated track of this traveler after a certain traveler.Complete after selection, track data will be presented in Main form scene.
The time shaft locomotive function of base map plane can make base map plane move on time-axis direction, to observe the residing locus of a certain moment track.In " time shaft moves " tool bar, selecting needs mobile base map plane graph layer and the moment moving to.After setting completes, the base map plane of choosing will move to corresponding moment position.
Tracing point time slice display function in scene, is convenient to user's observation analysis crowd's spatial distribution state and rule by the locus plotting of a certain all travelers of moment.In " time slice " tool bar, select base map plane graph layer, Data Date and section moment.After setting completes, choose all tracing points in moment will plotting on earth on plan.
Although embodiment of the present invention are open as above, but it is not restricted to listed utilization in instructions and embodiment, it can be applied to various applicable the field of the invention completely, for those skilled in the art, can easily realize other amendment, therefore do not deviating under the universal that claim and equivalency range limit, the present invention is not limited to specific details and illustrates here and the legend of describing.

Claims (8)

1. the stroke recognition methods based on trip track data, is characterized in that, comprises the following steps:
The speed of step 1, calculating tracing point;
Step 2, by multiple adjacent speed all the tracing point below threshold speed merge into a candidate stop place, wherein, the stop duration of described candidate stop place is that in described multiple tracing point, first tracing point arrives the time interval between last tracing point, the described multiple adjacent speed all tracing point below threshold speed is merged into a candidate stop place, be achieved in the following ways
(1) calculate successively the average coordinates (x of two adjacent track points in described candidate stop place (i, i+1), y (i, i+1)),
(2) calculate successively the time interval Δ t between described two adjacent track points (i, i+1), and the stop duration Stay Δ t of described candidate stop place between ratio wight (i, i+1)
(3) calculate described candidate stop place coordinate (Stay'x, Stay ' y):
Step 3, in the time that in the center of multiple candidates stop place and described multiple candidates stop place, the distance between any candidate stop place is less than distance threshold, and, when between zero hour of the stop duration of first candidate stop place in described multiple candidates stop place to the finish time of the stop duration of last candidate stop place interval greater than time threshold time, the center of described multiple candidates stop place is judged to be to dwell point, described step 3 is achieved in the following ways:
(1) using first candidate stop place in all candidates to be determined stop place as stopping sequence, wherein, described first candidate stop place is the center that stops sequence,
(2) in the time being positioned at distance that first candidate stop place at rear of described stop sequence arrives the center of described stop sequence and being less than distance threshold, described stop sequence is put into in first candidate stop place at the described rear that is positioned at described stop sequence, redefine the center of described stop sequence
(3) repeat (2), until when being positioned at distance that first candidate stop place at described stop sequence rear arrives the center of described stop sequence and being greater than distance threshold, when between zero hour of the stop duration of first candidate stop place in described stop sequence to the finish time of the stop duration of last candidate stop place interval greater than time threshold time, the center that stops sequence described in (2) is dwell point;
The time interval in step 4, described multiple candidates stop place between the zero hour to the finish time of the stop duration of last candidate stop place of the stop duration of first candidate stop place is the stop duration at described dwell point.
2. the stroke recognition methods based on trip track data as claimed in claim 1, is characterized in that, in described step 3, the center that stops sequence described in (2) is achieved in the following ways,
(1) calculating is positioned at the stop duration Stay ' of first candidate stop place at the rear of described stop sequence iratio wight between time interval Sq Δ t in Δ t and described stop sequence between the zero hour of the stop duration of first candidate stop place to the finish time of the stop duration of last candidate stop place i,
(2) calculate the centre coordinate (Sqx, Sqy) of described stop sequence:
Sq·x=wight i·Stay′i·x+(1-wight i)·Sq·x
Sq·y=wight i·Stay′i·y+(1-wight i)·Sq·y?。
3. the stroke recognition methods based on trip track data as claimed in claim 1, is characterized in that, also includes
Step 5, in described step 3, (3) when being positioned at distance that first candidate stop place at described stop sequence rear arrives the center of described stop sequence in and being greater than distance threshold, in the time that the time interval between the zero hour to the finish time of the stop duration of last candidate stop place of the stop duration of first candidate stop place in described stop sequence is less than time threshold, the all tracing points that stop described in (2) between zero hour to the finish time of the stop duration of last candidate stop place of stop duration of first candidate stop place in sequence are all judged to be transfer point, and in described stop sequence, all tracing points between finish time to the zero hour of the stop duration of first candidate stop place at described stop sequence rear of the stop duration of last candidate stop place are all judged to be transfer point.
4. the stroke recognition methods based on trip track data as claimed in claim 1, it is characterized in that, in described step 4, tracing points all between the zero hour of the stop duration of first candidate stop place in described multiple candidates stop place to the finish time of the stop duration of last candidate stop place is deleted.
5. the stroke recognition methods based on trip track data as claimed in claim 1, is characterized in that, in described step 1, calculates the average velocity of tracing point.
6. the stroke recognition methods based on trip track data as claimed in claim 5, is characterized in that, in described step 1, calculates the average velocity of tracing point, be achieved in the following ways,
At least one tracing point is chosen respectively at front and rear at current tracing point, calculate from first tracing point to the air line distance last tracing point, calculate first tracing point to the time interval between last tracing point, the average velocity of current tracing point is by obtaining to the time interval last tracing point from described first tracing point to the air line distance last tracing point divided by described first tracing point.
7. the stroke recognition methods based on trip track data as claimed in claim 5, is characterized in that, in described step 1, calculates the average velocity of tracing point, be achieved in the following ways,
At least one tracing point is chosen respectively at front and rear at current tracing point, calculate respectively two air line distances between adjacent track point, and calculate the air line distance sum between all two adjacent track points in selected tracing point, calculate first tracing point to the time interval between last tracing point, the average velocity of current tracing point obtains to the time interval between last tracing point divided by described first tracing point by the air line distance sum between all two adjacent track points in described selected tracing point.
8. the stroke recognition methods based on trip track data as claimed in claim 1, is characterized in that, described time threshold is 300 seconds, and described distance threshold is 1100 meters.
CN201210056545.5A 2012-03-06 2012-03-06 Travel track data-based stroke identification method Expired - Fee Related CN102607553B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201210056545.5A CN102607553B (en) 2012-03-06 2012-03-06 Travel track data-based stroke identification method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201210056545.5A CN102607553B (en) 2012-03-06 2012-03-06 Travel track data-based stroke identification method

Publications (2)

Publication Number Publication Date
CN102607553A CN102607553A (en) 2012-07-25
CN102607553B true CN102607553B (en) 2014-08-13

Family

ID=46525155

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201210056545.5A Expired - Fee Related CN102607553B (en) 2012-03-06 2012-03-06 Travel track data-based stroke identification method

Country Status (1)

Country Link
CN (1) CN102607553B (en)

Families Citing this family (77)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102879791B (en) * 2012-10-09 2014-05-14 潮州市创佳电子有限公司 System for sensing activity data of elder person based on Beidou positioning terminal
CN103150177B (en) * 2013-01-14 2019-05-24 北京百度网讯科技有限公司 A kind of methods, devices and systems updating public bus network data
CN103970805B (en) * 2013-02-05 2018-01-09 日电（中国）有限公司 Move Mode excavating equipment and method
CN103149577B (en) * 2013-02-28 2016-04-13 山东大学 The Combinated navigation method that " Big Dipper " navigation, GPS navigation and historical data merge
CN104050173B (en) * 2013-03-12 2020-11-03 百度在线网络技术（北京）有限公司 Method and system for constructing database for site semantic recognition
CN104515529A (en) * 2013-09-27 2015-04-15 高德软件有限公司 Real-scenery navigation method and navigation equipment
CN103697897A (en) * 2013-12-06 2014-04-02 成都亿盟恒信科技有限公司 Inflection point compensation of vehicle track information
CN103745083B (en) * 2013-12-11 2017-01-25 深圳先进技术研究院 Trajectory data cleaning method and device
CN104732756A (en) * 2013-12-24 2015-06-24 中兴通讯股份有限公司 Method for conducting public transportation planning by utilizing mobile communication data mining
CN104156489B (en) * 2014-08-29 2017-11-28 北京嘀嘀无限科技发展有限公司 The method that the resident point excavation of driver is carried out based on driver track
CN104239556B (en) * 2014-09-25 2017-07-28 西安理工大学 Adaptive trajectory predictions method based on Density Clustering
CN104463420B (en) * 2014-11-05 2017-11-21 上海携程商务有限公司 The order processing system and method for OTA websites
CN105608919B (en) * 2014-11-21 2019-04-09 杭州海康威视数字技术股份有限公司 The determination method and device of interchange of position
CN104537052B (en) * 2014-12-26 2017-06-30 西南交通大学 Traffic trip transfer point recognition methods based on wavelet analysis Study of modulus maximum algorithm
CN104504099B (en) * 2014-12-29 2018-02-16 北京交通大学 Traffic trip state cutting method based on location track
US9904932B2 (en) * 2014-12-29 2018-02-27 Google Llc Analyzing semantic places and related data from a plurality of location data reports
CN104767534B (en) * 2014-12-30 2017-07-25 中移全通系统集成有限公司 A kind of controllable track of vehicle point compression and storage method of error and system
CN104615881B (en) * 2015-01-30 2017-12-22 南京烽火星空通信发展有限公司 A kind of user's normality trajectory analysis method based on shift position application
CN104596507B (en) * 2015-02-09 2017-10-03 成都小步创想畅联科技有限公司 A kind of determination method of mobile terminal trip track
CN105243396A (en) * 2015-11-06 2016-01-13 百度在线网络技术（北京）有限公司 User position information generation method and device
CN105608505B (en) * 2015-12-22 2020-02-11 重庆邮电大学 Resident rail transit trip mode identification method based on mobile phone signaling data
CN105469435A (en) * 2016-01-20 2016-04-06 北京格灵深瞳信息技术有限公司 Track compression method and device
CN105740904B (en) * 2016-01-29 2019-10-11 东南大学 A kind of trip based on DBSCAN clustering algorithm and activity pattern recognition methods
CN106101999B (en) * 2016-05-27 2019-06-11 广州杰赛科技股份有限公司 A kind of recognition methods of user trajectory and device
CN106856498A (en) * 2016-06-27 2017-06-16 芮锶钶（上海）网络技术有限公司 A kind of method that owner's operating motor vehicles for determining mobile communication terminal are left
CN106227889A (en) * 2016-08-15 2016-12-14 华云科技有限公司 A kind of track dwell point analyzes extracting method
CN106384120B (en) * 2016-08-29 2019-08-23 深圳先进技术研究院 A kind of resident's activity pattern method for digging and device based on mobile phone location data
CN106570184B (en) * 2016-11-11 2020-08-14 同济大学 Method for extracting recreation-living contact data set from mobile phone signaling data
CN106776902A (en) * 2016-11-30 2017-05-31 北京锐安科技有限公司 The analysis method and device of path locus
CN106792514B (en) * 2016-11-30 2020-10-30 南京华苏科技有限公司 User position analysis method based on signaling data
CN106778857B (en) * 2016-12-09 2019-08-06 北京首都国际机场股份有限公司 The method obtained automatically to port flight baggage status
CN106772513A (en) * 2016-12-16 2017-05-31 中国航天系统工程有限公司 A kind of tourist security managing device and system based on the Big Dipper
CN108256596A (en) * 2016-12-29 2018-07-06 航天信息股份有限公司 Mobile device management method, apparatus and system in grain depot based on active label
CN107071721B (en) * 2016-12-31 2020-06-05 景致惠通工程咨询（武汉）有限公司 Stopover point extraction method based on mobile phone positioning data
CN108574933B (en) * 2017-03-07 2020-11-27 华为技术有限公司 User track recovery method and device
CN108692724A (en) * 2017-04-10 2018-10-23 中兴通讯股份有限公司 A kind of indoor navigation method and device
CN107328952B (en) * 2017-05-05 2019-06-18 四川大学 Flight path velocity stability evaluation scheme when a kind of no true value
CN108021625B (en) * 2017-11-21 2021-01-19 深圳广联赛讯股份有限公司 Vehicle abnormal gathering place monitoring method and system, and computer readable storage medium
CN108398702B (en) * 2017-12-12 2020-08-11 北京荣之联科技股份有限公司 Parking environment identification method and device
CN110031876A (en) * 2018-01-11 2019-07-19 中南大学 A kind of vehicle mounted guidance tracing point offset antidote based on Kalman filtering
CN108168562B (en) * 2018-01-31 2020-04-14 上海势航网络科技有限公司 Method for extracting stop point of positioning track
CN110118567B (en) * 2018-02-06 2021-07-20 北京嘀嘀无限科技发展有限公司 Travel mode recommendation method and device
CN108376415B (en) * 2018-02-13 2022-01-21 中国联合网络通信集团有限公司 Track filling method and device
CN108763553B (en) * 2018-06-01 2021-07-20 云南大学 Density-based dwell point identification method
CN109104694B (en) * 2018-06-26 2020-10-30 重庆市交通规划研究院 User stay position finding method and system based on mobile phone signaling
CN110738228B (en) * 2018-07-20 2023-05-02 菜鸟智能物流控股有限公司 Track processing method and device and electronic equipment
CN110807074A (en) * 2018-08-01 2020-02-18 山东华软金盾软件股份有限公司 Method for realizing dynamic track playing based on map JS technology
CN110826758B (en) * 2018-08-14 2023-10-13 上海零数众合信息科技有限公司 Stroke type determining method and device, storage medium and terminal
CN110873783A (en) * 2018-08-14 2020-03-10 上海能链众合科技有限公司 Carbon emission monitoring method and device, storage medium and terminal
CN112739984A (en) * 2018-09-13 2021-04-30 华为技术有限公司 Mobile phone multimodal position sensing
AU2018286582A1 (en) 2018-09-18 2020-04-02 Beijing Didi Infinity Technology And Development Co., Ltd. Artificial intelligent systems and methods for predicting traffic accident locations
CN109633716B (en) * 2018-12-10 2020-10-27 东南大学 GPS-based urban distribution vehicle travel chain and characteristic identification method and equipment thereof
CN109886724B (en) * 2018-12-29 2021-02-12 中南大学 Robust resident travel track identification method
CN111696343B (en) * 2019-03-12 2022-04-05 北京嘀嘀无限科技发展有限公司 Track data processing method and device
CN111770432B (en) * 2019-04-02 2022-07-15 北京三快在线科技有限公司 Method and device for identifying stop point, electronic equipment and storage medium
CN110213718A (en) * 2019-05-24 2019-09-06 北京小米移动软件有限公司 The method and device of perception terminal behavior
CN110290467B (en) * 2019-06-21 2020-08-04 清华大学 Method and device for acquiring stop point, business area service range and influence factor
CN110491157B (en) * 2019-07-23 2022-01-25 中山大学 Vehicle association method based on parking lot data and checkpoint data
CN110428621B (en) * 2019-07-30 2022-07-15 山东交通学院 Track data-based monitoring and early warning method for dangerous driving behavior of floating car
CN110428604B (en) * 2019-07-30 2022-04-22 山东交通学院 Taxi illegal parking monitoring and early warning method based on track and map data
CN110545522B (en) * 2019-08-13 2021-06-01 广州瀚信通信科技股份有限公司 User position and functional area identification method based on mobile big data
CN110689804B (en) * 2019-10-10 2022-05-17 百度在线网络技术（北京）有限公司 Method and apparatus for outputting information
CN111159582B (en) * 2019-12-20 2023-10-31 北京邮电大学 Method and device for processing track data of moving object
CN111046049B (en) * 2019-12-20 2021-10-26 西南交通大学 Truck GPS track data compression method
CN111190891B (en) * 2019-12-27 2023-07-25 武汉长江通信产业集团股份有限公司 Multi-semantic track data segment storage method
CN111340331B (en) * 2020-02-10 2023-11-14 泰华智慧产业集团股份有限公司 Analysis method and system for residence behavior of supervisor in city management work
CN111508228B (en) * 2020-04-01 2021-01-01 佛山市城市规划设计研究院 Method for acquiring public transport trip chain by using mobile phone GPS and electronic map data
CN111461077B (en) * 2020-05-12 2024-01-12 北京爱笔科技有限公司 Method and device for identifying movement track event
CN111625610A (en) * 2020-05-25 2020-09-04 重庆大学 Display method and system for generating track based on positioning points
CN111881242A (en) * 2020-07-28 2020-11-03 腾讯科技（深圳）有限公司 Basic semantic recognition method of track points and related equipment
CN112100408B (en) * 2020-08-24 2022-05-31 北京完美知识科技有限公司 Historical entity data display method, device and equipment
CN112468973B (en) * 2020-11-10 2023-06-27 恒安嘉新（北京）科技股份公司 Optimization method, device, equipment and medium for signaling positioning track
CN112835080B (en) * 2021-01-21 2024-03-19 成都路行通信息技术有限公司 Track repairing method and device for vehicle in stationary state and electronic equipment
CN113065064B (en) * 2021-03-24 2023-09-29 支付宝(杭州)信息技术有限公司 Information recommendation processing method, device, equipment and storage medium
CN114509076B (en) * 2022-02-16 2023-10-20 平安科技（深圳）有限公司 Method, device, equipment and storage medium for processing movement track data
CN114863715A (en) * 2022-05-05 2022-08-05 一汽解放汽车有限公司 Parking data determination method and device, electronic equipment and storage medium
CN117171605B (en) * 2023-11-03 2024-02-20 中国林业科学研究院森林生态环境与自然保护研究所(国家林业和草原局世界自然遗产保护研究中心) Migration bird track segmentation method based on GPS data

Family Cites Families (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
DE60037360T2 (en) * 1999-05-28 2008-12-04 Nippon Telegraph And Telephone Corp. Method and device for measuring the speed of vehicles with an image processing system
JP3867696B2 (en) * 2003-10-06 2007-01-10 住友電気工業株式会社 Moving means discriminating apparatus and method, and OD traffic volume calculating apparatus and method
JP2008146249A (en) * 2006-12-07 2008-06-26 Nippon Telegraph & Telephone West Corp Probe data analysis system
JP2008146248A (en) * 2006-12-07 2008-06-26 Nippon Telegraph & Telephone West Corp Probe data analysis system
CN100580735C (en) * 2008-09-25 2010-01-13 北京航天智通科技有限公司 Real-time dynamic traffic information processing method based on car detecting technique
US10657738B2 (en) * 2008-10-27 2020-05-19 International Business Machines Corporation Reconstructing an accident for a vehicle involved in the accident
US20100211308A1 (en) * 2009-02-19 2010-08-19 Microsoft Corporation Identifying interesting locations
US8275649B2 (en) * 2009-09-18 2012-09-25 Microsoft Corporation Mining life pattern based on location history
CN101739825A (en) * 2009-11-06 2010-06-16 吉林大学 GPS floating vehicle-based traffic data fault identification and recovery method
JP5471930B2 (en) * 2010-07-20 2014-04-16 株式会社デンソー Driving assistance device

Also Published As

Publication number Publication date
CN102607553A (en) 2012-07-25

Similar Documents

Publication Publication Date Title
CN102607553B (en) Travel track data-based stroke identification method
CN110298500B (en) Urban traffic track data set generation method based on taxi data and urban road network
CN107767659B (en) Shared bicycle attraction amount and occurrence amount prediction method based on ARIMA model
Krueger et al. Visual analysis of movement behavior using web data for context enrichment
Furletti et al. Inferring human activities from GPS tracks
CN103038605B (en) The position probing of passing through for barrier, timetable and traveling time in numerical map are estimated
Ahas et al. Mobile positioning in space–time behaviour studies: social positioning method experiments in Estonia
CN109859495A (en) A method of overall travel speed is obtained based on RFID data
CN110716935A (en) Track data analysis and visualization method and system based on online taxi appointment travel
CN105806351A (en) Road information prediction method and device
CN110837973B (en) Human trip selection information mining method based on traffic trip data
Bosurgi et al. Pavement condition information modelling in an I-BIM environment
Keler et al. Detecting vehicle traffic patterns in urban environments using taxi trajectory intersection points
US20230245002A1 (en) Stationary Classifier for Geographic Route Trace Data
Bierlaire et al. Modeling route choice behavior from smartphone GPS data
Yuan et al. Recognition of functional areas based on call detail records and point of interest data
Chen et al. Enabling smart urban services with gps trajectory data
CN115412857A (en) Resident travel information prediction method
CN109800903A (en) A kind of profit route planning method based on taxi track data
CN110276563A (en) A kind of mode of transportation transfer Activity recognition method based on supporting vector machine model
CN110413855A (en) A kind of region entrance Dynamic Extraction method based on taxi drop-off point
Guillermo et al. Graph Database-modelled Public Transportation Data for Geographic Insight Web Application
CN102881182A (en) Traffic information display method and device
Simkowitz Transportation applications of geographic information systems
CN114666738A (en) Territorial space planning method and system based on mobile phone signaling

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
C14 Grant of patent or utility model
GR01 Patent grant
CF01 Termination of patent right due to non-payment of annual fee

Granted publication date: 20140813

Termination date: 20170306

CF01 Termination of patent right due to non-payment of annual fee