CN108985488A - The method predicted to individual trip purpose - Google Patents
The method predicted to individual trip purpose Download PDFInfo
- Publication number
- CN108985488A CN108985488A CN201810573420.7A CN201810573420A CN108985488A CN 108985488 A CN108985488 A CN 108985488A CN 201810573420 A CN201810573420 A CN 201810573420A CN 108985488 A CN108985488 A CN 108985488A
- Authority
- CN
- China
- Prior art keywords
- grid
- trajectory
- sub
- theme
- obtains
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 31
- 238000007667 floating Methods 0.000 claims abstract description 25
- 238000004458 analytical method Methods 0.000 claims abstract description 7
- 238000005070 sampling Methods 0.000 claims description 9
- 238000005457 optimization Methods 0.000 claims description 6
- 239000011159 matrix material Substances 0.000 claims description 4
- 230000000295 complement effect Effects 0.000 claims 1
- 238000004445 quantitative analysis Methods 0.000 abstract description 5
- 230000006399 behavior Effects 0.000 description 4
- 238000010586 diagram Methods 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 2
- 238000004513 sizing Methods 0.000 description 2
- 230000002123 temporal effect Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000010267 cellular communication Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000004590 computer program Methods 0.000 description 1
- 235000013399 edible fruits Nutrition 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 230000001105 regulatory effect Effects 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 238000012163 sequencing technique Methods 0.000 description 1
- 239000013589 supplement Substances 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G06Q50/40—
Abstract
A kind of method that the present invention predicts with providing individual trip purpose, method includes: to carry out grid dividing to geographic range corresponding to wheelpath, and grid ID is arranged；The wheelpath is segmented, a plurality of sub-trajectory is obtained；It matches each tracing point and grid of the sub-trajectory, obtains grid ID collection corresponding with each sub-trajectory；The number of default first theme carries out modeling analysis to each sub-trajectory and its corresponding grid ID collection according to LDA topic model, obtains the first prediction model of corresponding first theme；It is calculated according to current driving data of first prediction model to a Floating Car, obtains the probability value of corresponding each first theme.The present invention can carry out quantitative analysis based on trip intention of a large amount of wheelpath data to driver, and with predicting next trip purpose, can be good at the planning for being applied to traffic path, or carry out the prediction etc. of short-term traffic flow.
Description
Technical field
The present invention relates to field of traffic, the method predicted with particularly relating to individual trip purpose.
Background technique
Reiseziel and next section are the key that driving behaviors, concerning the safety and convenience of people's trip.
Therefore, the valuable letters such as the trip intention of driver, next section how are excavated from the traffic track data of magnanimity
There are demands for breath.
Summary of the invention
The technical problems to be solved by the present invention are: a kind of method predicted while individual trip purpose is provided, it can be to a
Body trip intention is predicted.
In order to solve the above-mentioned technical problem, the technical solution adopted by the present invention are as follows:
A kind of method predicted of individual trip purpose ground, comprising:
Travelling data according to a Floating Car obtains corresponding wheelpath；
Grid dividing is carried out to geographic range corresponding to the wheelpath, and grid ID is set；
The wheelpath is segmented according to preset time difference, obtains a plurality of sub-trajectory；
It matches each tracing point and grid of the sub-trajectory, obtains grid ID collection corresponding with each sub-trajectory；
The number of default first theme, according to LDA topic model to each sub-trajectory and its corresponding grid ID collection into
Row modeling analysis obtains the first prediction model of corresponding first theme；
It is calculated according to current driving data of first prediction model to a Floating Car, it is each to obtain correspondence
The probability value of first theme.
The beneficial effects of the present invention are: based on providing a kind of travelling data by driver, according to LDA subject pattern
Be analyzed and processed and method that the trip purpose of car owner is predicted and the computer program for realizing this method.This
Invention can find travel event behavior subject pattern from a large amount of wheelpath data, so to the trip intention of driver into
Row quantitative analysis, and with predicting next section or trip purpose, it can not only be applied to the rule of traffic path well
It draws, also can be carried out the prediction etc. of short-term traffic flow.
Detailed description of the invention
The flow diagram for the method that Fig. 1 predicts for a kind of individual trip purpose of the present invention；
Fig. 2 is the flow diagram of the embodiment of the present invention.
Specific embodiment
To explain the technical content, the achieved purpose and the effect of the present invention in detail, below in conjunction with embodiment and cooperate attached
Figure is explained.
The most critical design of the present invention is: travel event behavior theme mould can be found from a large amount of wheelpath data
Formula, and then quantitative analysis is carried out to the trip intention of driver, and with predicting next section or trip purpose.
Embodiment
Referring to Fig.1 and 2, a kind of method predicted the present embodiment provides individual trip purpose, can be realized and be based on
The current travelling data of driver predicts its intention of going on a journey.
Referring particularly to Fig. 2, the method for the present embodiment the following steps are included:
S1: the travelling data according to a Floating Car obtains corresponding wheelpath；
Specifically, the board information terminal in the Floating Car driven using driver is during floating vehicle travelling with one
Mode by wireless communication is sent after the travelling data of preset fixed time period T acquisition Floating Car, and to be sent to back-end data flat
Platform.Wherein, the travelling data includes car number u, position l, speed v, time t；Driving is collected according to time sequencing
Data sequence, and background data center is sent by mobile cellular communication technology by the travelling data sequence of acquisition.
After background data center receives the travelling data of a Floating Car, basic information (such as vehicle of the corresponding Floating Car
Number etc. and the unique corresponding identification information of the Floating Car) it stores into database.Pass through the Floating Car that will constantly collect
Travelling data store to be located at distal end background data center, realize board information terminal memory space optimization, mitigate it
Storage burden.
When the request predicted when the next running section or trip purpose for receiving the corresponding Floating Car, Ke Yiyou
Background service center, which calls directly, to be corresponded to the data of the Floating Car and is analyzed and processed in its database, obtain result；Certainly,
The request can be initiated by any one terminal, and after the related data that background service center obtains Floating Car, in terminal local
Result is obtained after being analyzed and processed.The selection of the actuating station of prediction process can carry out flexible configuration according to different usage scenarios
(such as having may be selected when good network state is supported may be selected eventually when backstage executes, terminal has superior performance configuration support
End executes), better meet actual demand.Preferably, in order to mitigate the burden of terminal local, while rear Taiwan investment is made full use of
Source, actuating station of the default choice by background service center as prediction process.
In the following, will be unfolded to carry out specifically so that travelling data of the background data center to the Floating Car is handled as an example
It is bright；If selection is handled by terminal locally, in terminal after getting travelling data from the background, using terminal as executing master
Body is also able to achieve by same processing mode.
After the travelling data of the corresponding Floating Car of background data center acquisition request promoter, corresponding driving is obtained
Track can obtain the driving in corresponding geographical location after being ranked up scattered travelling data according to chronological order
Track.
S2: grid dividing is carried out to geographic range corresponding to the wheelpath, and grid ID is set；
Background data center confirms the active geographic range of the Floating Car, the concrete foundation wheelpath according to wheelpath
The geographical location information of record is confined；
Then grid dividing is carried out to identified geographic range, and corresponding ID is arranged to each grid.Wherein, it draws
The sizing grid divided is determined according to the specific tracing point of each of composition wheelpath.Since each tracing point records pair
The geographical location information answered preferably is made in order to ensure the subsequent accuracy handled based on gridding information the analysis of wheelpath
The mode that each tracing point can correspond to a grid carries out grid dividing.It therefore, can be with each specific longitude and latitude of direct basis
The geographical coordinate of information carries out grid dividing, and the preferably described sizing grid is 100m*100m, i.e., each corresponding determination of grid
The geographical location of latitude and longitude information.Preferably, its ID is arranged in the corresponding geographical location information of each grid of direct basis.
Such as: get the history travelling data that a car number u is 5205278；According to travelling data acquisition and its
Corresponding wheelpath；Then its scope of activities, i.e. geographic range are determined according to the wheelpath, preferably chooses and is greater than the driving
The a certain range of geographic range in track；Grid dividing subsequently is carried out to identified geographic range, and grid ID is set, is established
Grid model corresponding with the wheelpath.Specifically, grid spacing is 100 meters, each tracing point of wheelpath is extracted
Geographical location information, i.e. latitude and longitude information combine and determine the corresponding grid ID of the tracing point.Such as: Fuzhou City center
(119.300023754,26.080006881), it is 0.001 that longitude and latitude precision, which is first arranged, obtains (119.300,26.080), is led to
Calculating 119.300*1000*100000+26.080*1000=11930026080 is crossed, the ID of the position is obtained are as follows:
11930026080。
In this step, after confining corresponding geographic range according to wheelpath, only to the range carry out grid dividing and
Grid ID is set, rather than entire map is all subjected to grid dividing, data volume to be dealt with can be substantially reduced, only for
Valid data are analyzed and processed, and realize the optimization of system processing mode, and then promote the realization speed of the method for the present embodiment.
S3: the wheelpath is segmented according to preset time difference, obtains a plurality of sub-trajectory；
Specifically, can be according to the time difference (such as Δ t > 30min) of two neighboring travelling data as segmentation foundation, to whole
Track data is segmented, and a plurality of sub-trajectory m is formed.For example, the temporal information of travelling data 1 is 2015-07-14 18:
The temporal information of 26:27, travelling data 2 are 2015-07-14 18:26:37, and two data time, which subtract each other, can be obtained the time difference 10
Second；Then it is greater than 30 minutes (Δ t > 30min) as foundation is segmented, to whole track data according to the time difference of adjacent two data
It is segmented, forms 100 sub-trajectories.
The purpose being segmented to wheelpath is modeled according to LDA topic model to the sub-trajectory of each item to be subsequent
Analysis provides data basis, i.e. the calculating of the present embodiment is carried out for a plurality of sub-trajectory in driver's certain time interval
Quantitative analysis processing as a result, therefore, acquired calculated result can more preferably, more accurately embody driver and correspond to the above-mentioned time
Trip intention in interval, can realize more accurate prediction.
S4: matching each tracing point and grid of the sub-trajectory, obtains each corresponding grid of sub-line wheel paths
ID collection；
By the way that each tracing point of each sub-trajectory is matched to grid division according to its geographical location information recorded
Geographic range in, thus obtain the corresponding grid collection of each sub-trajectory and grid ID set.
Briefly, i.e., the corresponding each tracing point of each sub-trajectory is obtained first；Then each rail is matched
Mark point and grid；Obtain the corresponding grid ID collection of each sub-trajectory；
For example, corresponding to its position l (longitude and latitude recorded in travelling data according to a tracing point in a wherein sub-trajectory
Spend information) it is (119.300,26.080), which is matched in the body of a map or chart of grid division on corresponding position,
Then the corresponding grid ID in the position in the map is obtained.
By the step, each tracing point and the matching of grid map of the sub-trajectory of each item are realized, and respectively corresponding
Grid ID collection acquisition.
It in a specific embodiment, further include process of data preprocessing, i.e., before next step S5 after the step
Optimize the corresponding relationship of sub-trajectory and grid map；
Specifically, including data interpolating and data deduplication step, the S41 and S42 in Fig. 2 are respectively corresponded:
(1) data interpolating
Since travelling data acquisition periodically carries out, track data through processing does not have sparsity.
In order to which the track for travelling driver is more preferably attached, interpolation one by one is carried out to sub- track data using linear interpolation techniques and is mended
It fills, that is, ensures that the corresponding grid of the tracing point of every sub-trajectory is continuous.
Specifically, can by successively traversing each sub-trajectory, using linear interpolation techniques to currently traverse this
Each tracing point of sub-trajectory carries out interpolation one by one and supplements, and then enables every sub-trajectory corresponding with continuous grid, inserts
Value treated each sub-trajectory is optimization sub-trajectory；Finally obtain the corresponding grid ID collection of each optimization sub-trajectory.
For example, two neighboring tracing point 1,2 is belonging respectively to 11930026081 and of grid ID in a sub-trajectory
11930026083, i.e. intermediate mesh ID11930026082 does not have data, then being just inserted into data to grid ID, that is, is inserted into one
A tracing point forms it into a continuous track, which is the sub-trajectory after optimizing.
(2) data deduplication
Duplicate removal is carried out to adjacent repeated grid, according to time sequence, if adjacent data point (tracing point) belongs to a net
Lattice, i.e., adjacent repeated grid then only retain a grid ID；
Specifically, by each sub-trajectory of traversal, each tracing point pair corresponding to this sub-trajectory currently traversed
The grid ID answered is judged, carries out duplicate removal to adjacent repeated grid, such as: continuous three data, grid ID are
11930026080,11930026080,11930026081, then it is left two data 11930026080 after duplicate removal,
11930026081。
By data interpolating and data deduplication step, be able to ascend be subsequently used for modeling basic data accuracy and
Integrality, and then be conducive to final calculation result, that is, the accuracy for the destination predicted.
S5: the number of default first theme, according to LDA topic model to each sub-trajectory and its corresponding grid ID collection
Modeling analysis is carried out, the first prediction model of corresponding first theme is obtained.
Application method of the LDA topic model in traffic trip field is Trajectory (track) -> topic (main
Topic) -> OD (traffic start-stop section).
Specifically, the step may include following sub-step:
S51: parameter configuration；
The number of first theme K, the i.e. quantity of model calculating the first theme to be obtained are set according to driving behavior, such as
K=100；The parameter alpha of model is set, β is set as 0.0001, the i.e. setting of computational accuracy.
S52: random initializtion；
Grid ID i.e. corresponding to each tracing point of every sub-trajectory of wheelpath, it is random to assign topic (the
One theme number).
Specifically, every sub-trajectory will be traversed, net corresponding to each tracing point to this sub-trajectory currently traversed
Lattice ID assigns the first theme topic number at random, until each grid ID of the corresponding grid ID collection of all sub-trajectories is
Random assignment is completed, and the random assigned value of each grid ID is different.
The range of the number is corresponding with the number of first theme, i.e., determines according to K, if K is 100, range can
To be 0-99；
For example, being 11930026080 random imparting topic numbers 99 to grid ID, each sub-trajectory is successively traversed, to each
The corresponding grid ID of each tracing point of sub-trajectory assigns the number of a 0-99 at random, and each number is different.
S53:Gibbs Sampling sampling；
Each sub-trajectory is rescaned, re-starts and adopts according to grid ID of the Gibbs Sampling formula to its each tracing point
Sample updates topic number；The Gibbs Sampling formula is
Wherein, the V is corpus, stores all grid ID；T indicates t-th of grid ID in corpus；M is sub-trajectory
The total number of point；K indicates k-th of first themes；W indicates the corresponding grid ID collection of the m articles sub-trajectory；ziIt is implicit variable, table
Show the corresponding first theme topic number of i-th of grid ID；α, β respectively indicate the parameter of the first prediction model, are set as
0.0001；Indicate the number that each grid ID occurs in the first theme k；Indicate time that the first theme k occurs in sub-trajectory m
Number；Indicate the grid ID that i is designated as under removal.
Specifically, each sub-trajectory will be traversed again, each grid ID is updated using Gibbs Sampling sampling
Corresponding topic number, such as: the topic number of grid ID11930026080 random initializtion is 99, utilizes Gibbs
Sampling samples to obtain a topic number 1, then the topic number of grid ID11930026080 is updated to 1.
S54: repeating step S53, until convergence；The number of general iteration is 20000 times.
S55: counting each word frequency, calculates and obtains topic-ID frequency matrix in track, i.e., the LDA theme of corresponding first theme
Model, referred to as the first prediction model；
Specifically, being exactly to count the frequency and first theme that the first theme topic number occurs in each sub-trajectory
The frequency that each grid ID occurs in topic number；Then statistical result is substituted into the topic-ID frequency matrix of LDA topic model
It is calculated, obtains the first prediction model for corresponding to first theme.
For example, the total grid number 1000 of statistics, the number 100 that grid ID11930026080 occurs in theme 1, in theme 1
The total degree 10000 that all grid ID occur, the number 10000 that theme 1 occurs in sub-trajectory 1, all grid ID in sub-trajectory 1
The total degree 10000 of appearance, utilizes formulaIt is calculated: 1 grid ID11930026080's of theme
Parameter (100+0.0001)/(10000+0.0001*1000)=0.010；Above-mentioned formula can refer to " LDA mathematics Eight Diagrams " chapter 6
The related content of LDA text modeling, wherein the document in the corresponding text of sub-trajectory；Word in the corresponding text of grid ID.
And so on, all grid ID of all themes are traversed, its parameter is calculated, obtains topic-ID parameter matrix.
What the step was calculated is the first prediction model of corresponding first theme, and the quantity of theme is larger, close theme
It is more, therefore the calculation amount based on current driving data in the first prediction model may be larger, calculates more time-consuming, calculating knot
The precision of fruit may have realization to be hoisted, but having no effect on method, i.e. first prediction model still can apply to this
It predicts the destination of driver.Advanced optimize that treated the second prediction model, Neng Goujin is provided by following step simultaneously
One step promotes prediction treatment effeciency and accuracy.
S6: being handled by self-adaption cluster, similar first theme is gathered for one kind, the X number of topics for obtaining and determining
Corresponding prediction model.
Specifically, carrying out clustering according to the first theme to first prediction model, corresponding second theme is obtained
Second prediction model；
For example, clustered to 100 the first themes obtained above, same or similar track is gathered for one kind, finally
Obtain 10 determining numbers of topics, i.e. second theme；And matched with map, obtain 10 places: engineering college, Drum Hill,
Wanda, Forest Park, Baolong, East Street mouth, agricultural university, permanently happy airport, Pingtan, Yongtai；The place illustrated is located at Fujian Province Foochow
City.
S7: calculating the current driving data of a Floating Car according to prediction model, obtains corresponding each theme
Probability value, and accordingly carry out destination recommendation.
Specifically, only providing the first prediction model if considering based on cost etc., then the progress of the first prediction model is directly based upon
It calculates, driver is recommended into the corresponding geographical location of maximum first theme of probability value；, but it is preferable that pre- by described second
It surveys model to calculate the current driving data of a Floating Car, obtains the probability value of corresponding each second theme；
Then it is carried out the corresponding geographical location of the maximum second theme of probability value as next destination of the driver of the Floating Car
Recommend.
For example, using the trip purpose built up prediction model, i.e.,This 10 themes are calculated in parameter constant
Probability, choose the theme of maximum probability as the trip purpose of prediction: the driving is currently in the main road Pu Shang
(119.264928131,26.037414520), current travelling data with being sent to the trip purpose built up prediction model
Background data center or terminal, be calculated the next destination of the driving and its corresponding probability value: engineering college 50%,
Drum Hill 8%, Wanda 15%, Forest Park 0.2%, Baolong 10%, East Street mouthful 10%, agricultural university 5%, permanently happy airport 1%, Pingtan
0.5%, Yongtai 0.3%；The work institute that selection probability is 50% is as the trip purpose predicted.
Since the present embodiment is directed to the prediction result that each sub-trajectory carries out quantitative analysis processing, predicted
Corresponding destination is next more short-circuit journey destination, to guarantee the accuracy of prediction result, while prediction result also more accords with
The actual demand for closing driver can be realized intelligence, precisely, rapidly carry out in advance to the destination of lower a road section of driver
It surveys, is conducive to the planning (setting of navigation destination is rapidly completed in the destination of such as direct basis prediction) of travel route, especially
It is the planning in travelling, helps the clear present running route of driver；Further, traffic control can also be applied to well,
By the prediction result of next driving path to driving vehicles all in traffic route, the prediction of short-term traffic flow is realized, with
Just the wagon flow situation for timely grasping urban traffic road, is regulated and controled and is dredged in time before there is traffic congestion, facilitated
Alleviate urban traffic pressure.
In conclusion a kind of method predicted of individual trip purpose provided by the invention ground, it can be to individual trip intention
It is predicted；Not only accuracy is high；And it is high-efficient；The present invention can be good at applying to the trip intention to driver personal
It is predicted, is conducive to trip planning；It can also apply to traffic control well, by driving vehicles all in traffic route
Next driving path prediction result, realize the prediction of short-term traffic flow.
The above description is only an embodiment of the present invention, is not intended to limit the scope of the invention, all to utilize this hair
Equivalents made by bright specification and accompanying drawing content are applied directly or indirectly in relevant technical field, similarly include
In scope of patent protection of the invention.
Claims (8)
1. a kind of method predicted to individual trip purpose characterized by comprising
Travelling data according to a Floating Car obtains corresponding wheelpath；
Grid dividing is carried out to geographic range corresponding to the wheelpath, and grid ID is set；
The wheelpath is segmented according to preset time difference, obtains a plurality of sub-trajectory；
It matches each tracing point and grid of the sub-trajectory, obtains grid ID collection corresponding with each sub-trajectory；
The number of default first theme, builds each sub-trajectory and its corresponding grid ID collection according to LDA topic model
Mould analysis obtains the first prediction model of corresponding first theme；
It calculates, obtains corresponding each described according to current driving data of first prediction model to a Floating Car
The probability value of first theme.
2. the method predicted to individual trip purpose as described in claim 1, which is characterized in that described pre- according to described first
It surveys model to calculate the current driving data of a Floating Car, obtains the probability value of corresponding each first theme,
Specifically:
Clustering is carried out according to the first theme to first prediction model, obtains the second prediction mould of corresponding second theme
Type；
It calculates, obtains corresponding each described according to current driving data of second prediction model to a Floating Car
The probability value of second theme；
Recommend the maximum second theme of probability value.
3. the method predicted to individual trip purpose as described in claim 1, which is characterized in that the matching sub-trajectory
Each tracing point and grid, obtain grid ID collection corresponding with each sub-trajectory, specifically:
Obtain the corresponding each tracing point of each sub-trajectory；
Match each tracing point and grid；
Obtain the corresponding grid ID collection of each sub-trajectory；
Each sub-trajectory is successively traversed, is inserted one by one using each tracing point of the linear interpolation method to the sub-trajectory currently traversed
Value complement is filled, and each optimization sub-trajectory corresponding with continuous grid is obtained；
Obtain the corresponding grid ID collection of each optimization sub-trajectory.
4. the method predicted to individual trip purpose as described in claim 1, which is characterized in that default first theme
Number carries out modeling analysis to each sub-trajectory and its corresponding grid ID collection according to LDA topic model, obtains described in corresponding to
First prediction model of the first theme, specifically:
The number K of default first theme；
Each sub-trajectory is traversed, assigns mono- the first theme of grid ID corresponding to each tracing point of each sub-trajectory at random
Topic number, the range of the number are corresponding with the number of first theme；
Each tracing point is traversed again using Gibbs Sampling formula, updates grid ID corresponding to each tracing point
The first theme topic number；
Traversal and the update again are repeated, until convergence；
Count each net in the frequency of occurrences and the first theme topic number that the first theme topic is numbered in each sub-trajectory
The frequency that lattice ID occurs, and the topic-ID frequency matrix for being substituted into LDA topic model is calculated, and obtains corresponding to described the
First prediction model of one theme.
5. the method predicted to individual trip purpose as claimed in claim 4, which is characterized in that the Gibbs Sampling
Formula is
Wherein, the V is corpus, stores all grid ID；T indicates t-th of grid ID in corpus；M is son
The total number of tracing point；K indicates k-th of first themes；W indicates the corresponding grid ID collection of the m articles sub-trajectory；ziIt is implicit change
Amount indicates the corresponding first theme topic number of i-th of grid ID；α, β respectively indicate the parameter of the first prediction model, setting
It is 0.0001；Indicate the number that each grid ID occurs in the first theme k；Indicate the first theme k appearance in sub-trajectory m
Number；Indicate the grid ID that i is designated as under removal.
6. the method predicted to individual trip purpose as claimed in claim 5, which is characterized in that the topic-ID frequency square
Battle array be
7. the method predicted to individual trip purpose as described in claim 1, which is characterized in that the travelling data includes floating
Motor-car number, time point information, geographical location information and vehicle speed information.
8. the method predicted to individual trip purpose as described in claim 1, which is characterized in that according to the corresponding longitude and latitude of grid
It spends information and grid ID is set.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810573420.7A CN108985488A (en) | 2018-06-06 | 2018-06-06 | The method predicted to individual trip purpose |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810573420.7A CN108985488A (en) | 2018-06-06 | 2018-06-06 | The method predicted to individual trip purpose |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108985488A true CN108985488A (en) | 2018-12-11 |
Family
ID=64540764
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810573420.7A Pending CN108985488A (en) | 2018-06-06 | 2018-06-06 | The method predicted to individual trip purpose |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108985488A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110160539A (en) * | 2019-05-28 | 2019-08-23 | 北京百度网讯科技有限公司 | Map-matching method, calculates equipment and medium at device |
| CN111929715A (en) * | 2020-06-28 | 2020-11-13 | 杭州云起智慧校园科技有限公司 | Positioning method, device and equipment for school badge and storage medium |
| CN113159105A (en) * | 2021-02-26 | 2021-07-23 | 北京科技大学 | Unsupervised driving behavior pattern recognition method and data acquisition monitoring system |
| CN113486719A (en) * | 2021-06-08 | 2021-10-08 | 南京邮电大学 | Vehicle destination prediction method, vehicle destination prediction device, computer equipment and storage medium |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104834679A (en) * | 2015-04-14 | 2015-08-12 | 苏州大学 | Representation and inquiry method of behavior track and device therefor |
| CN107861957A (en) * | 2016-09-22 | 2018-03-30 | 杭州海康威视数字技术股份有限公司 | A kind of data analysing method and device |
| CN108072378A (en) * | 2016-11-15 | 2018-05-25 | 中国移动通信有限公司研究院 | A kind of method and device for predicting destination |
-
2018
- 2018-06-06 CN CN201810573420.7A patent/CN108985488A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104834679A (en) * | 2015-04-14 | 2015-08-12 | 苏州大学 | Representation and inquiry method of behavior track and device therefor |
| CN107861957A (en) * | 2016-09-22 | 2018-03-30 | 杭州海康威视数字技术股份有限公司 | A kind of data analysing method and device |
| CN108072378A (en) * | 2016-11-15 | 2018-05-25 | 中国移动通信有限公司研究院 | A kind of method and device for predicting destination |
Non-Patent Citations (2)
| Title |
|---|
| 徐广根: "基于LDA 主题模型的用户电信轨迹恢复算法", 《HTTP://KNS.CNKI.NET/KCMS/DETAIL/51.1196.TP.20180424.1022.010.HTML》 * |
| 蔡文学: "基于LDA 的用户轨迹分析", 《计算机应用与软件》 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110160539A (en) * | 2019-05-28 | 2019-08-23 | 北京百度网讯科技有限公司 | Map-matching method, calculates equipment and medium at device |
| CN111929715A (en) * | 2020-06-28 | 2020-11-13 | 杭州云起智慧校园科技有限公司 | Positioning method, device and equipment for school badge and storage medium |
| CN113159105A (en) * | 2021-02-26 | 2021-07-23 | 北京科技大学 | Unsupervised driving behavior pattern recognition method and data acquisition monitoring system |
| CN113159105B (en) * | 2021-02-26 | 2023-08-08 | 北京科技大学 | Driving behavior unsupervised mode identification method and data acquisition monitoring system |
| CN113486719A (en) * | 2021-06-08 | 2021-10-08 | 南京邮电大学 | Vehicle destination prediction method, vehicle destination prediction device, computer equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108197739B (en) | Urban rail transit passenger flow prediction method | |
| CN108985488A (en) | The method predicted to individual trip purpose | |
| CN107103754B (en) | Road traffic condition prediction method and system | |
| CN104731963B (en) | Recommend method and system in a kind of gridding path based on car networking | |
| CN108256553B (en) | Construction method and device for double-layer path of vehicle-mounted unmanned aerial vehicle | |
| CN110517482B (en) | Short-term traffic flow prediction method based on 3D convolutional neural network | |
| CN103593430B (en) | Clustering method based on mobile object spatiotemporal information trajectory subsections | |
| CN110264709A (en) | The prediction technique of the magnitude of traffic flow of road based on figure convolutional network | |
| CN109670277A (en) | A kind of travel time prediction method based on multi-modal data fusion and multiple-model integration | |
| CN106710208A (en) | Traffic state acquisition method and device | |
| CN110599767A (en) | Long-term and short-term prediction method based on network taxi appointment travel demands | |
| CN109214462A (en) | A kind of spatial data on-line talking method based on distributed incremental DBSCAN algorithm | |
| CN103731916B (en) | A kind of customer location forecasting system and method based on wireless network | |
| CN109686091B (en) | Traffic flow filling algorithm based on multi-source data fusion | |
| CN108446470A (en) | Medical facilities analysis method of reachability based on track of vehicle data and population distribution | |
| CN107544516A (en) | Automated driving system and method based on relative entropy depth against intensified learning | |
| JP7226855B2 (en) | Calculation method, system and storage medium for train passing through tunnel | |
| CN110716935A (en) | Track data analysis and visualization method and system based on online taxi appointment travel | |
| CN112633602B (en) | Traffic congestion index prediction method and device based on GIS map information | |
| CN111144281B (en) | Urban rail transit OD passenger flow estimation method based on machine learning | |
| CN106899306A (en) | A kind of track of vehicle line data compression method of holding moving characteristic | |
| CN106875734A (en) | A kind of method and device for pushing parking route | |
| CN112419711B (en) | Closed parking lot parking demand prediction method based on improved GMDH algorithm | |
| CN113342029B (en) | Maximum sensor data acquisition path planning method and system based on unmanned aerial vehicle cluster | |
| CN109064750A (en) | Urban road network traffic estimation method and system |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20181211 |