CN108256560A - A kind of park recognition methods based on space-time cluster - Google Patents
A kind of park recognition methods based on space-time cluster Download PDFInfo
- Publication number
- CN108256560A CN108256560A CN201711448160.2A CN201711448160A CN108256560A CN 108256560 A CN108256560 A CN 108256560A CN 201711448160 A CN201711448160 A CN 201711448160A CN 108256560 A CN108256560 A CN 108256560A
- Authority
- CN
- China
- Prior art keywords
- park
- time
- point
- label
- trip
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2413—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on distances to training or reference patterns
- G06F18/24147—Distances to closest patterns, e.g. nearest neighbour classification
Landscapes
- Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Traffic Control Systems (AREA)
Abstract
The invention discloses a kind of park recognition methods based on space-time cluster, include the following steps：Based on the GPS track data of smart mobile phone acquisition individual activity, space time information is extracted；Any tracing point nearest k point on a timeline is searched for, the core point in track is determined by distance parameter Eps and minimal amount threshold value MinPts；Core point continuous in time forms initial cluster, is checked since the cluster of density maximum, the adjacent cluster neighbouring to space-time merges, so as to be parked；Non-core point continuous in time forms initial trip, is checked since time earliest trip, if the time interval with latter trip is less than the minimum duration threshold value of park, the two is merged, and correct pseudo- park as trip.The present invention can rapidly and accurately identify the park in individual trip GPS track, lay the foundation for further identification trip mode with trip purpose, and technical support is provided for long-term, extensive, passive type Urban Residential Trip investigation.
Description
Technical field
The invention belongs to the cluster analysis of traffic data excavation applications more particularly to time series data and individual trip rails
Park identification in mark.
Background technology
With the quick universal and development of smart mobile phone, accurate positioning function, abundant sensor module go out for individual
Acquisition provides hardware condition in real time for row track.In face of track data of largely going on a journey, analysis individual behavior feature, identification activity
Pattern becomes the main bugbear of data service urban transportation.It is identified and parked based on individual GPS track data, be to judge OD, speculate
The premise of trip mode and trip purpose works.Current correlative study mainly according under static state velocity characteristic, movement side
It is subject to regular judgement to feature, and with reference to road network.And feature is gathered according to tracing point and is parked from Density Clustering angle recognition, it is existing
Technology focuses primarily upon DBSCAN methods.
DBSCAN sets Eps neighborhoods and minimum points two parameters of MinPts, reachable with density based on core point
For classifying mode, constantly extension is connected to form a cluster outward.However time series a little is not accounted for, for space length phase
Closely, easily wrong merger is one kind to the remote point cluster of time interval, and can not be accurate for situations such as coincidence path, trip in short-term
Identification park.Specifically, its algorithm specifically has the disadvantage that when handling smart mobile phone GPS mobile datas：Processing is when extremely big
When measuring tracing point, distance matrix EMS memory occupation is big, leads to that program cannot respond to or arithmetic speed is slow；Parameter Eps and MinPts is quick
Perceptual high, generalization ability is poor；Trip often, path overlap high not easy to identify of degree, especially for the portion of track overlapping
Pavement branch sections or intersection are because a large amount of tracing points gather, so as to which false judgment is park.
Traffic trip origin and destination recognition methods of the number of patent application 201611195129.3 based on Spact-time clustering algorithm,
Core concept is that the space length extended in DBSCAN is time-space matrix, control parameter △ T, Eps, MinPts is set, not mark
The point of note starts to search for core point to surrounding, so as to obtain initial cluster；Merge cluster by the threshold values of 600 seconds or 500m, gone out
Row origin and destination.But in the case of trip complexity or bad signal quality, merge threshold value and not necessarily adapt to；Merge order to exist
Randomness has an impact recognition result；The puppet of the stop in short-term park of wrong identification is not rejected, trip number result will be higher.
Therefore the clustering method for considering Spatial dimensionality is explored, adapts to complicated trip or data, efficiently and accurately identification trip
Park in track, it is significant to urban transportation data mining, trip pattern-recognition etc..
Invention content
In order to solve the above technical problems, the purpose of the present invention is to provide provide a kind of Density Clustering for considering time series
Method, for handling the individual trip GPS data of smart mobile phone acquisition and providing park identification operation interface and result to the user
Visualization.
It is shown with visualization interface.
The present invention is achieved by the following technical solutions：
1st, a kind of park recognition methods based on space-time cluster, includes the following steps：
S1, the GPS track data based on smart mobile phone acquisition individual activity, from the GPS track extracting data space-time
Information obtains time series data set；
S2, the park identified in individual trip GPS track is handled the time series data set using clustering algorithm,
And export park recognition result；
S3, the park recognition result is evaluated；
S4, park recognition user interface visualization display.
Further, in step S1, the GPS track data are acquired with identical sample frequency；The space time information packet
Include time, longitude and latitude information；The time series data collection is combined into is converted to plane coordinate value by the longitude and latitude
Data acquisition system D={ P afterwardsi, i=1,2 ..., M, wherein M are the number of tracing point, and D is the data acquisition system of M tracing point, sub
Collect PiFor the data acquisition system of i-th of tracing point, Pi=(ti, xi, yi), wherein, ti, xi, yiThe time of respectively i-th tracing point,
Abscissa value, ordinate value.
Further, in step S2, the clustering algorithm processing further comprises the steps：
S21, setting clustering algorithm parameter, including search length k, time neighbouring threshold value I, spatial neighbor threshold value Eps,
Core point differentiates the minimum duration DU of minimal amount MinPts, park；
S22, label is initialized, all tracing point initial markers are Label=∞ in the data acquisition system D, and park is initial
Number is ClustID=1；
S23, search core point, for tracing point P each in the data acquisition system Di, determine the tracing point PiOn time
K closest point, and each point in the k point and the tracing point P are calculated respectivelyiSpace length, if the sky
Between distance be less than the spatial neighbor threshold value Eps the N that counts outepsDifferentiate minimal amount MinPts more than the core point, then
It will point PiLabeled as core point, and label Label=-1 is changed, otherwise keep label Label=∞；
S24, park merges and label, is core point set { C by the core point merger continuous in timej}；Search for core
The core point set C of heart point set density DS maximumsm, calculate the core point set CmWith adjacent point set { Cm-1,Cm+1Time-space matrix；
If the adjacent residence time interval LTWith the adjacent park space length LDIn threshold range, then merge two point sets,
Otherwise by point set CmLabeled as park, update label Label=ClustID and cluster number ClustID=ClustID+1, and search
The core point set of the next density maximum of rope, until all core point sets complete label；
S25, go out performing check, for the point of Label=∞ in the data acquisition system D, merger is if continuous in time
Row set { TRn}；If the time interval of adjacent trip is less than the minimum duration DU of the park, merge trip, update pseudo- park
Interior point marked as Label=0；
Time Continuous and point merger that is adjacent, having same label are same class by S26, output recognition result, and park is marked
Number it is more than 0, goes out line label equal to 0；Calculate park, trip starts and end time, and calculates park center position coordinates, output
Recognition result.
Further, in step S21, described search length k refers to for any tracing point Pi, at a distance of most on the time
K near point { PSi,...,PTi, wherein, PSiFor starting point, PTiFor terminal, the label (S, T) of the starting and terminal point is expressed as：
The time is neighbouring, refers to and parks A to Mr. Yuj, adjacent park { A on the timej-1,Aj+1, if meeting latter stop
Adjacent residence time interval L between start time and previous park finish timeTThe threshold value I neighbouring less than the time；
The spatial neighbor refers to and parks A to Mr. Yuj, adjacent park { A on the timej-1,Aj+1, if meeting latter stop
Adjacent park space length L between center and previous park centerDLess than the threshold value of the spatial neighbor
Eps。
Further, the adjacent residence time interval LTIt is expressed as：
Wherein, TjsIt is carved at the beginning of for j-th of park, Tj,EFinish time for j-th of park；
The adjacent park space length LDIt is expressed as：
Wherein, Xj,YjCenter position coordinates for j-th of park.
Further, in step S24, further comprise following steps：
S241, by Label=-1 and the core point merger continuous in time it is core point set Cj, calculate each institute
State core point set CjInitial time TjS, end time TjE, center position coordinates (Xj,Yj), wherein Xj, YjRespectively described core
Heart point set CjIn all tracing points are horizontal, average value of ordinate；
S242, core point set density DS is defined, is the core point number NP and the core that a core point is concentrated
Maximum space distance d between point each pointmaxRatio, DS=NP/dmax；
The core point set CjIn, DSj=NPj/djmax；
S243, search obtain the core point set C of the core point set density DS maximumsm, wherein the core point collection CmRise
Begin moment TmS, end time TmE, center position coordinates, (Xm,Ym), wherein Xm, YmRespectively described core point set CmMiddle institute's rail
Mark point is horizontal, the average value of ordinate；
Calculate the core point set CmWith adjacent point set { Cm-1,Cm+1Time-space matrix；If the point set Cm-1With the core
Heart point set CmThe adjacent residence time interval LT=Tjs-Tj-1,E＜ I and the adjacent park space lengthThen merge point set { Cm-1,Cm, and change CmInitial time TmS=
Tm,S-1；If Cm+1Meet the adjacent residence time interval LT=Tj+1,s-Tj,E＜ I and the adjacent park space lengthThen merge point set { Cm,Cm+1, and change CmEnd time TmE=
Tm,E+1；If { Cm-1,Cm+1It is unsatisfactory for merging condition, it is determined that CmTo park, and update label Label=ClustID；
Step S244：Cluster number ClustID=ClustID+1 is updated, searches for next density maximum and Label=-1
Core point set repeats step S243, until all core point sets complete label.
Further, the step S25 further comprises following steps：
S251, for the point of Label=∞ and Time Continuous in the data acquisition system D, form out row set { TRn, meter
Calculate the initial time T of rownSWith end time TnE；
S252, for go on a journey TRn, calculate and subsequent period trip TRn+1Time interval LT=Tn+1,S-Tn,EIf LT＜
DU illustrates TRnWith TRn+1Between park duration be unsatisfactory for minimum duration requirement, then merge trip { TRn,TRn+1, and more
Change end time TnE=Tn,E+1；
S253：By merging go out line label and between pseudo- park label be changed to Label=0, repeat step S252, directly
To all trip test endings.
Further, in step S3, park recognition result evaluation index includes,
(1) trip number consistency TripNumFit
If the trip times N of predictionPTWith true trip times NTTIt is equal, then it is 1, is otherwise 0；
(2) trip start/stop time difference TimeDiff
The prediction initial time t of all tripspsWith end time tpeRespectively with the initial time t of log recordingtsWith termination
Moment tteMean difference；
(3) park duration difference DUDiff
The prediction duration DU of all parkspkWith the park duration DU of log recordingtkMean difference；
(4) moving position deviation distance DistP
The prediction center P of all parkspcWith actual position PtcAverage departure distance；
Wherein, NTATo park number.
Further, park recognition user interface visualization display, inputs the cluster in user oriented interface
Algorithm parameter, display stops identification figure and the start-stop position residence time being parked in GPS data and the park recognition result
Evaluation index.
Further, the GPS track data sampling frequency is 1Hz.
The minimum duration DU of park, whether the cluster for judging identification is park, and threshold value is defaulted as 120s.
Present invention identification core point, searches for k point nearest on its time shaft, when Eps models since the point of track first
Interior points are enclosed when reaching MinPts, then judge that the point is core point, depend primarily on k and Eps.
The merging of the present invention calculates previous park finish time and latter park start time difference and previous park
Center and latter park center position, if being respectively smaller than parameter I and Eps, merge.
Pseudo- park filtering of the invention, due to being stopped in track there may be accidental, as signalized intersections encounter red light, peak
Situations such as section congestion, bus stop are waited filters it by setting minimum park duration, can obtain more accurately identifying
As a result.
Compared with prior art, the present invention has the advantages that：
The present invention compensates for the defects of DBSCAN processing time sequence datas, by the way that setting time is neighbouring, spatial neighbor threshold
Situations such as value, counterweight combining diameter, trip in short-term, can accurately identify, and with better generalization ability.
The present invention judges core point to search for the point near k, and initial cluster is generated from the angle of Spatiotemporal Aggregation；Define cluster
Density reflection Spatial Agglomeration degree, and with the merging order of this determinant, so as to avoid randomness, reduce and merge mistake；
Threshold value is merged by setting, is adapted to the situation that trip is complicated or signal quality is bad；Minimum park duration, filtering are considered simultaneously
Puppet park.
The present invention can rapidly and accurately identify the park in individual trip GPS track, for further identification trip mode
It lays the foundation with trip purpose, technical support is provided for long-term, extensive, passive type Urban Residential Trip investigation.
Description of the drawings
Fig. 1 is the present invention is based on the park recognition methods flow charts that space-time clusters.
Fig. 2 identifies case effect track I recognition result schematic diagrames for present invention park, and heavy line is park, and fine line is
Trip.
Fig. 3 identifies case effect track II recognition result schematic diagrames for present invention park, and heavy line is park, and fine line is
Trip.
Fig. 4 is present invention park identification visualization display figure.
Specific embodiment
Elaborate below to the embodiment of the present invention, the present embodiment with the technical scheme is that according to development,
Give detailed embodiment and specific operating process.
Case study on implementation：
The full-time trip GPS track of two smart mobile phone acquisitions of selection, track I features are long for residence time, put in space
On more dissipate, record the time be 6:37:02~20:06:53, wherein including 3 parks and 2 trips；Track II features are
It goes on a journey often, and path overlaps degree height, the record time is 7:04:13~18:44:05, wherein including 6 parks and 5 times
Trip.
Arrange parameter：Search length k=61, time proximity threshold I=30s, spatial neighbor threshold value Eps=30m, core point
Differentiate the minimum duration DU=120s of minimal amount MinPts=30, park.
The trip daily record of table 1
To the trip daily record GPS track data of table 1, according to shown in Fig. 1, the present invention is based on the park recognition methods of space-time cluster
Flow chart, processing identifies the park in track I and track II, and exports park recognition result.
For track I recognition results as shown in Fig. 2, heavy line is park, fine line is trip.
For track II recognition results as shown in figure 3, heavy line is park, fine line is trip.
According to evaluation index, the park recognition result of track I and track II are evaluated, as table 2 parks identification and evaluation
It is shown.
Table 2 parks identification and evaluation
Present invention park identification can visualize display, and such as Fig. 4 inputs clustering algorithm parameter in display interface, and display stops
Identification figure and the start-stop position residence time being parked in GPS data and park recognition result evaluation index.
Above example be the application preferred embodiment, those of ordinary skill in the art can also on this basis into
The various transformation of row or improvement, under the premise of the total design of the application is not departed from, these transformation or improvement should all belong to this Shen
Within the scope of please being claimed.
Claims (10)
1. a kind of park recognition methods based on space-time cluster, which is characterized in that include the following steps：
S1, the GPS track data based on smart mobile phone acquisition individual activity, from the GPS track extracting data space time information,
Obtain time series data set；
S2, the park identified in individual trip GPS track is handled the time series data set using clustering algorithm, and defeated
Go out to park recognition result；
S3, the park recognition result is evaluated；
S4, park recognition user interface visualization display.
2. a kind of park recognition methods based on space-time cluster according to claim 1, which is characterized in that in step S1,
The GPS track data are acquired with identical sample frequency；The space time information includes time, longitude and latitude information；It is described
Time series data collection is combined into the data acquisition system D={ P after the longitude and latitude to be converted to plane coordinate valuei, i=1,
The number of 2 ..., M, wherein M for tracing point, data acquisition systems of the D for M tracing point, subset PiData set for i-th of tracing point
It closes, Pi=(ti, xi, yi), wherein, ti, xi, yiThe time of respectively i-th tracing point, abscissa value, ordinate value.
3. a kind of park recognition methods based on space-time cluster according to claim 2, which is characterized in that in step S2,
The clustering algorithm processing further comprises the steps：
S21, setting clustering algorithm parameter, neighbouring threshold value I, the threshold value Eps of spatial neighbor, core including search length k, time
Point differentiates the minimum duration DU of minimal amount MinPts, park；
S22, label is initialized, all tracing point initial markers are Label=∞ in the data acquisition system D, park initial number
For ClustID=1；
S23, search core point, for tracing point P each in the data acquisition system Di, determine the tracing point PiIt is most adjacent on time
K near point, and each point in the k point and the tracing point P are calculated respectivelyiSpace length, if the space away from
The N that counts out from the threshold value Eps less than the spatial neighborepsDifferentiate minimal amount MinPts more than the core point, then by point
PiLabeled as core point, and label Label=-1 is changed, otherwise keep label Label=∞；
S24, park merges and label, is core point set { C by the core point merger continuous in timej}；Search for core point set
The core point set C of density DS maximumsm, calculate the core point set CmWith adjacent point set { Cm-1,Cm+1Time-space matrix；It is if described
Adjacent residence time interval LTWith the adjacent park space length LDIn threshold range, then merge two point sets, otherwise will
Point set CmLabeled as park, update label Label=ClustID and cluster number ClustID=ClustID+1, and search for next
The core point set of a density maximum, until all core point sets complete label；
S25, go out performing check, for the point of Label=∞ in the data acquisition system D, merger is trip collection if continuous in time
Close { TRn}；If the time interval of adjacent trip is less than the minimum duration DU of the park, merge trip, update point in pseudo- park
Marked as Label=0；
Time Continuous and point merger that is adjacent, having same label are same class by S26, output recognition result, and park label is big
In 0, go out line label equal to 0；Calculate park, trip starts and end time, and calculates park center position coordinates, output identification
As a result.
4. a kind of park recognition methods based on space-time cluster according to claim 3, which is characterized in that in step S21,
Described search length k refers to for any tracing point Pi, at a distance of k nearest point { P on the timeSi,...,PTi, wherein, PSi
For starting point, PTiFor terminal, the label (S, T) of the starting and terminal point is expressed as：
The time is neighbouring, refers to and parks A to Mr. Yuj, adjacent park { A on the timej-1,Aj+1, start if meeting latter park
Adjacent residence time interval L between moment and previous park finish timeTThe threshold value I neighbouring less than the time；
The spatial neighbor refers to and parks A to Mr. Yuj, adjacent park { A on the timej-1,Aj+1, if meeting latter park center
Adjacent park space length L between position and previous park centerDLess than the threshold value Eps of the spatial neighbor.
5. a kind of park recognition methods based on space-time cluster according to claim 4, which is characterized in that described adjacent to stop
L in time intervalTIt is expressed as：
Wherein, TjsIt is carved at the beginning of for j-th of park, Tj,EFinish time for j-th of park；
The adjacent park space length LDIt is expressed as：
Wherein, Xj,YjCenter position coordinates for j-th of park.
6. a kind of park recognition methods based on space-time cluster according to claim 3, which is characterized in that in step S24,
Further comprise following steps：
S241, by Label=-1 and the core point merger continuous in time it is core point set Cj, calculate each described core
Heart point set CjInitial time TjS, end time TjE, center position coordinates (Xj,Yj), wherein Xj, YjRespectively described core point
Collect CjIn all tracing points are horizontal, average value of ordinate；
S242, core point set density DS is defined, the core point number NP and the core point concentrated for a core point is each
Maximum space distance d between pointmaxRatio, DS=NP/dmax；
The core point set CjIn, DSj=NPj/djmax；
S243, search obtain the core point set C of the core point set density DS maximumsm, wherein the core point collection CmStarting when
Carve TmS, end time TmE, center position coordinates, (Xm,Ym), wherein Xm, YmRespectively described core point set CmIn all tracing points
Horizontal, ordinate average value；
Calculate the core point set CmWith adjacent point set { Cm-1,Cm+1Time-space matrix；If the point set Cm-1With the core point
Collect CmThe adjacent residence time interval LT=Tjs-Tj-1,E＜ I and the adjacent park space lengthThen merge point set { Cm-1,Cm, and change CmInitial time TmS=
Tm,S-1；If Cm+1Meet the adjacent residence time interval LT=Tj+1,s-Tj,E＜ I and the adjacent park space lengthThen merge point set { Cm,Cm+1, and change CmEnd time TmE=
Tm,E+1；If { Cm-1,Cm+1It is unsatisfactory for merging condition, it is determined that CmTo park, and update label Label=ClustID；
Step S244：Cluster number ClustID=ClustID+1 is updated, searches for next density maximum and the core of Label=-1
Point set repeats step S243, until all core point sets complete label.
A kind of 7. park recognition methods based on space-time cluster according to claim 3, which is characterized in that the step
S25 further comprises following steps：
S251, for the point of Label=∞ and Time Continuous in the data acquisition system D, form out row set { TRn, calculate trip
Initial time TnSWith end time TnE；
S252, for go on a journey TRn, calculate and subsequent period trip TRn+1Time interval LT=Tn+1,S-Tn,EIf LT＜ DU, say
Bright TRnWith TRn+1Between park duration be unsatisfactory for minimum duration requirement, then merge trip { TRn,TRn+1, and change end
Only moment TnE=Tn,E+1；
S253：By merging go out line label and between pseudo- park label be changed to Label=0, repeat step S252, Zhi Daosuo
There is trip test ending.
8. a kind of park recognition methods based on space-time cluster according to claim 1, which is characterized in that in step S3,
Park recognition result evaluation index includes,
(1) trip number consistency TripNumFit
If the trip times N of predictionPTWith true trip times NTTIt is equal, then it is 1, is otherwise 0；
(2) trip start/stop time difference TimeDiff
The prediction initial time t of all tripspsWith end time tpeRespectively with the initial time t of log recordingtsWith end time
tteMean difference；
(3) park duration difference DUDiff
The prediction duration DU of all parkspkWith the park duration DU of log recordingtkMean difference；
(4) moving position deviation distance DistP
The prediction center P of all parkspcWith actual position PtcAverage departure distance；
Wherein, NTATo park number.
9. a kind of park recognition methods based on space-time cluster according to claim 8, which is characterized in that park identification is used
Family interface visualization is shown, the clustering algorithm parameter is inputted in user oriented interface, and display stops identification figure and park
Start-stop position residence time and the park recognition result evaluation index in GPS data.
A kind of 10. park recognition methods based on space-time cluster according to claim 2, which is characterized in that the GPS rails
Mark data sampling frequency is 1Hz.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711448160.2A CN108256560B (en) | 2017-12-27 | 2017-12-27 | Parking identification method based on space-time clustering |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711448160.2A CN108256560B (en) | 2017-12-27 | 2017-12-27 | Parking identification method based on space-time clustering |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108256560A true CN108256560A (en) | 2018-07-06 |
| CN108256560B CN108256560B (en) | 2021-05-04 |
Family
ID=62724228
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711448160.2A Expired - Fee Related CN108256560B (en) | 2017-12-27 | 2017-12-27 | Parking identification method based on space-time clustering |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108256560B (en) |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108763553A (en) * | 2018-06-01 | 2018-11-06 | 云南大学 | Dwell point recognition methods based on density |
| CN109684384A (en) * | 2018-12-29 | 2019-04-26 | 太原科技大学 | A kind of track data space-time density Analysis System and its analysis method |
| CN109711447A (en) * | 2018-12-19 | 2019-05-03 | 武大吉奥信息技术有限公司 | A kind of special population event early warning and monitoring method and device |
| CN109948701A (en) * | 2019-03-19 | 2019-06-28 | 太原科技大学 | A kind of data clustering method based on space-time relationship between track |
| CN111159582A (en) * | 2019-12-20 | 2020-05-15 | 北京邮电大学 | Method and device for processing track data of moving object |
| CN111198929A (en) * | 2020-01-06 | 2020-05-26 | 中山大学 | Automatic processing method and system for fusing GPS track and activity log data |
| CN112085268A (en) * | 2020-08-31 | 2020-12-15 | 北京百度网讯科技有限公司 | Method, device and equipment for measuring and calculating resident travel information and readable storage medium |
| CN113469600A (en) * | 2020-03-31 | 2021-10-01 | 北京三快在线科技有限公司 | Travel track segmentation method and device, storage medium and electronic equipment |
| CN115964545A (en) * | 2023-03-16 | 2023-04-14 | 四川国蓝中天环境科技集团有限公司 | Method for deducing pollution point location based on slag transport vehicle track point |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2011128921A1 (en) * | 2010-04-15 | 2011-10-20 | Neptuny S.R.L. | Automated service time estimation method for it system resources |
| CN106384120A (en) * | 2016-08-29 | 2017-02-08 | 深圳先进技术研究院 | Mobile phone positioning data based resident activity pattern mining method and device |
| CN106600960A (en) * | 2016-12-22 | 2017-04-26 | 西南交通大学 | Traffic travel origin and destination identification method based on space-time clustering analysis algorithm |
| CN106971546A (en) * | 2017-05-18 | 2017-07-21 | 重庆大学 | Section bus permeability method of estimation based on bus GPS data |
| CN107305590A (en) * | 2017-06-14 | 2017-10-31 | 北京市交通信息中心 | A kind of urban transportation trip characteristicses based on mobile phone signaling data determine method |
-
2017
- 2017-12-27 CN CN201711448160.2A patent/CN108256560B/en not_active Expired - Fee Related
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2011128921A1 (en) * | 2010-04-15 | 2011-10-20 | Neptuny S.R.L. | Automated service time estimation method for it system resources |
| CN106384120A (en) * | 2016-08-29 | 2017-02-08 | 深圳先进技术研究院 | Mobile phone positioning data based resident activity pattern mining method and device |
| CN106600960A (en) * | 2016-12-22 | 2017-04-26 | 西南交通大学 | Traffic travel origin and destination identification method based on space-time clustering analysis algorithm |
| CN106971546A (en) * | 2017-05-18 | 2017-07-21 | 重庆大学 | Section bus permeability method of estimation based on bus GPS data |
| CN107305590A (en) * | 2017-06-14 | 2017-10-31 | 北京市交通信息中心 | A kind of urban transportation trip characteristicses based on mobile phone signaling data determine method |
Non-Patent Citations (3)
| Title |
|---|
| DERYA BIRANT 等: "ST-DBSCAN: An algorithm for clustering spatial–temporal data", 《DATA & KNOWLEDGE ENGINEERING 》 * |
| PALMA 等: "A clustering-based approach for discovering interesting places in trajectories", 《ACM SYMPOSIUM ON APPLIED COMPUTING》 * |
| 王忠民 等: "基于改进DBSCAN的移动用户兴趣点提取方法", 《西安邮电大学学报》 * |
Cited By (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108763553B (en) * | 2018-06-01 | 2021-07-20 | 云南大学 | Density-based dwell point identification method |
| CN108763553A (en) * | 2018-06-01 | 2018-11-06 | 云南大学 | Dwell point recognition methods based on density |
| CN109711447A (en) * | 2018-12-19 | 2019-05-03 | 武大吉奥信息技术有限公司 | A kind of special population event early warning and monitoring method and device |
| CN109684384A (en) * | 2018-12-29 | 2019-04-26 | 太原科技大学 | A kind of track data space-time density Analysis System and its analysis method |
| CN109948701B (en) * | 2019-03-19 | 2022-08-16 | 太原科技大学 | Data clustering method based on space-time correlation among tracks |
| CN109948701A (en) * | 2019-03-19 | 2019-06-28 | 太原科技大学 | A kind of data clustering method based on space-time relationship between track |
| CN111159582A (en) * | 2019-12-20 | 2020-05-15 | 北京邮电大学 | Method and device for processing track data of moving object |
| CN111159582B (en) * | 2019-12-20 | 2023-10-31 | 北京邮电大学 | Method and device for processing track data of moving object |
| CN111198929A (en) * | 2020-01-06 | 2020-05-26 | 中山大学 | Automatic processing method and system for fusing GPS track and activity log data |
| CN111198929B (en) * | 2020-01-06 | 2023-03-28 | 中山大学 | Automatic processing method and system for fusing GPS track and activity log data |
| CN113469600A (en) * | 2020-03-31 | 2021-10-01 | 北京三快在线科技有限公司 | Travel track segmentation method and device, storage medium and electronic equipment |
| CN112085268A (en) * | 2020-08-31 | 2020-12-15 | 北京百度网讯科技有限公司 | Method, device and equipment for measuring and calculating resident travel information and readable storage medium |
| CN112085268B (en) * | 2020-08-31 | 2024-03-05 | 北京百度网讯科技有限公司 | Method, device, equipment and readable storage medium for measuring and calculating resident trip information |
| CN115964545A (en) * | 2023-03-16 | 2023-04-14 | 四川国蓝中天环境科技集团有限公司 | Method for deducing pollution point location based on slag transport vehicle track point |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108256560B (en) | 2021-05-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108256560A (en) | A kind of park recognition methods based on space-time cluster | |
| Fang et al. | FTPG: A fine-grained traffic prediction method with graph attention network using big trace data | |
| CN106097755A (en) | For identifying the method parked place and/or vacate place | |
| EP2603770B1 (en) | Parking lot detection using probe data | |
| CN109084794B (en) | Path planning method | |
| CN110176139A (en) | A kind of congestion in road identification method for visualizing based on DBSCAN+ | |
| CN111081047A (en) | Accurate intelligent parking management method and management system based on photoelectric image processing | |
| CN111915887A (en) | Integration and processing system and method based on multi-source heterogeneous traffic data | |
| CN103500504A (en) | Urban road traffic parameter estimation and road situation discrimination method based on RFID (Radio Frequency Identification) data and application system | |
| CN101739824A (en) | Data fusion technology-based traffic condition estimation method | |
| CN110807917A (en) | Highway intelligent security system based on cloud calculates | |
| CN109035783A (en) | A kind of virtual networks missing section automatic identifying method based on public transport GPS track | |
| CN109903574A (en) | The acquisition methods and device of crossing traffic information | |
| CN114363316A (en) | Intelligent networking monitoring and supervision system for cross-regional road infrastructure | |
| CN101131796A (en) | Road traffic parameter checking device and method thereof | |
| CN113361825A (en) | Early warning method and system for trampling accident | |
| CN113868492A (en) | Visual OD (origin-destination) analysis method based on electric police and checkpoint data and application | |
| CN113434619A (en) | 4g intelligent highway traffic road condition monitoring system | |
| CN109489679A (en) | A kind of arrival time calculation method in guidance path | |
| CN111159625A (en) | Urban green wave coordination effect evaluation system | |
| CN105489010B (en) | A kind of through street journey time reliability monitoring analysis system and method | |
| Zhou et al. | Method for judging parking status based on yolov2 target detection algorithm | |
| CN116823572B (en) | Population flow data acquisition method and device and computer readable storage medium | |
| CN111444286B (en) | Long-distance traffic node relevance mining method based on trajectory data | |
| CN103700259B (en) | Real-time traffic method of discrimination |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20210504 Termination date: 20211227 |