CN109215347A - A kind of traffic data quality control method based on crowdsourcing track data - Google Patents
A kind of traffic data quality control method based on crowdsourcing track data Download PDFInfo
- Publication number
- CN109215347A CN109215347A CN201811227001.4A CN201811227001A CN109215347A CN 109215347 A CN109215347 A CN 109215347A CN 201811227001 A CN201811227001 A CN 201811227001A CN 109215347 A CN109215347 A CN 109215347A
- Authority
- CN
- China
- Prior art keywords
- data
- track
- point
- crowdsourcing
- limit
- 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
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/01—Detecting movement of traffic to be counted or controlled
- G08G1/0104—Measuring and analyzing of parameters relative to traffic conditions
- G08G1/0125—Traffic data processing
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/01—Detecting movement of traffic to be counted or controlled
- G08G1/0104—Measuring and analyzing of parameters relative to traffic conditions
- G08G1/0137—Measuring and analyzing of parameters relative to traffic conditions for specific applications
Landscapes
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Traffic Control Systems (AREA)
Abstract
This patent discloses a kind of method of quality control of crowdsourcing track data, which comprises Step 1: data cleansing step;Step 2: abnormal track data identification and processing step;Step 3: data set track pre-generatmg and set-up procedure;Step 4: complete trajectory presents and updates historical data base.Fault data present in traffic data of the present invention for crowdsourcing track data, missing data propose solution, it is ensured that the integrality of track data provides accurate data supporting for traffic programme, traffic forecast.
Description
Technical field
The present invention relates to intelligent transportation fields, more particularly to a kind of traffic data quality control based on crowdsourcing track data
Method processed.
Background technique
With the high speed development of wireless sensor technology, the information processing technology, development of Mobile Internet technology and satellite positioning tech,
The acquisition modes of traffic data are gradually diversified, detect from traditional earth coil, infrared video based on mobile terminal
Car networking, traffic data tends to magnanimity and diversification.In recent years, the proposition of crowdsourcing provides one kind for traffic data collection
The mode of new low cost, various difference mobile units and the sporadic data on sensor, the mobile device work with ordinary user
Perception data, the communication of vehicle vehicle and the interaction number between bus or train route, people's vehicle collected by way of communication for basic sension unit
All it is the source of crowdsourcing data according to grade, can be used for the data acquisition of Real-time Traffic Information.Crowdsourcing track data is exactly in space-time
Under situation, based on Small objects groups such as pedestrian, taxis, obtained by the data collection to these mobile object motion processes
Data information, have position, time, the speed etc. of sequencing including collecting.But due to the positioning means taken of individual are different,
The factors such as unstable, the external environment interference of communication link, the track data for frequently resulting in acquisition exists in precision and attribute to be owed
It lacks.Crowdsourcing track data is more huge compared to traditional section detection data data volume, city area coverage is bigger, while it
It is required that crowdsourcing participant provides it in the space time information of real physical world, it is related to huge privacy of user data.But crowdsourcing
Track of vehicle data contained again road grid traffic information, including vehicle space-time position, speed, following distance abundant etc. specifically number
According to the reliability of accuracy and decision to Traffic Stream of Road prediction result has an important influence on, thus the present invention proposes one kind
Traffic data quality control method based on crowdsourcing track data, it is intended to by certain data processing method to the numerous of acquisition
The traffic data of original crowdsourcing track data carries out quality control, so that it is guaranteed that the accuracy of data, consistency and integrality,
The traffic data quality of highway network is improved, provides complete, accurate data supporting for highway network traffic programme and management.
It is mostly to the road traffic flow data detected based on traditional sensors, including traffic in existing country's patent
The parameters such as amount, average speed, roadway occupancy carry out quality evaluation and control, and control method is substantially for according to traffic three elements
Between relationship carry out data exception identification and repair, but the traffic data that detects of traditional section sensor with based on user's individual
Both crowdsourcing track datas data attribute it is different, the also difference, therefore existing traffic data quality in process demand
Control method is not particularly suited for the processing of crowdsourcing track data.
Summary of the invention
To solve the above-mentioned problems, the present invention proposes a kind of traffic data quality controlling party based on crowdsourcing track data
Method.Data quality control method is mainly by data cleansing, data-privacy, abnormal track data identification and processing, data set rail
Mark is generated to be constituted with four aspects of adjustment.
Data cleansing refers to extracts key feature in numerous data, rejects irrelevant information and is only ground with regard to critical field
Study carefully, and the input format of data, the integrality of data field, the integrality of data bulk after extraction are detected.With vehicle
For crowdsourcing track data, field mainly includes vehicle ID, order ID, time point, geographical location, section number, lane volume
Number etc., carry out data transmission according to the format of standard, by taking the time as an example, input format is xxxx-xx-xx xx:xx:xx, ground
Manage position is indicated with conventional longitude and latitude, number of the integrality of data bulk should acquire under data volume size and standard time interval
It is compared according to amount size.
Data-privacy is that anonymous processing is carried out for the data content containing individual information, in order to keep the accurate of data
Property, thus to the elaborate position information for the crowdsourcing participant for including in the crowdsourcing track data based on space-time without fuzzy place
Reason only has the data of obvious user information to carry out privacy processing with regard to vehicle ID, user mobile phone condition code etc..
Abnormal track data identification refers to whether detection data shortage of data occurs, whether is more than parameter threshold with processing
Deng.The setting of parameter threshold is according to category of roads, service level, traffic control and management level, vehicle running characteristics, road ring
The series of factors such as border characteristic and practical experience are formulated, and the factor of threshold decision mainly includes threshold speed and vehicle driving road
Crowdsourcing track data is compared with the upper lower threshold value of setting parameter, fills out after rejecting fault data to data by diameter threshold value
It studies for a second time courses one has flunked again, forms complete data set.
It is that complete crowdsourcing track data will be matched with road network after handling that data set track, which is generated with adjustment, extraction phase
It with vehicle ID, is ranked up according to time order and function, longitude and latitude is accurately matched with road network position, adjacent data point is connected and is formed
Running track carries out track correction, denoising according to the track of generation, to obtain smooth complete accurately vehicle operation rail
Mark.
Specific step is as follows for a kind of method of quality control of crowdsourcing track data proposed by the present invention:
Step 1, data cleansing: since crowdsourcing Data Data amount is huge, the acquisition information content is abundant, is related to and traffic number more
According to irrelevant contents, it is therefore desirable to crowdsourcing track data be carried out key message extraction, be entered into database according to the format of agreement
In, it carries out the correction of data standard time point and detects acquired data attribute format, field contents, data set size, step
Suddenly are as follows:
Step 1.1, the data that will acquire carry out key feature extraction, reject redundant field, memory space needed for reducing, and
Data inputting format is set according to agreement, the attribute integrality of inspection data: sets each field before data inputting database
The data are labeled as 0 if format mismatching by standard typing format, judge it for data fault;
Step 1.2 successively differentiates each critical field of data, detects whether there are certain field contents to be empty or lattice
Field contents are that messy code can not identify in the matched situation of formula, are 0 by the incomplete data markers of field contents, judge it for number
According to failure;
Step 1.3, the correction of standard time point and detection data collection size: by 24 hours one day according to equipment detection cycle T
Be divided into period standard time for time interval, since crowdsourcing track data acquisition time interval is shorter, data acquisition frequently and
There are small difference for the average time interval of actual path detection data, not exclusively exactly equal to preset acquisition week
Phase, real time point can slowly be detached from the standard time point system set up with the drift of time, thus need to click through the time
Row amendment carries out the judgement of data set size again.Tentative standard time point is t', and preset measurement period is T, allows to adopt
Integrate time delay as t, statistics gatherer data period is that the data in (t'+t-T, t'+t) carry out time point amendment, and principle is as follows:
(1) if the detection data for falling into the period is recorded as 0, this standard time point data missing, by the point data mark
It is denoted as 0, judges it for shortage of data;
(2) if the detection data for falling into the period is recorded as 1, the numerical value of this standard time point is unique detection data
The numerical value of record;
(3) there is a plurality of, the segment limit when numerical value of this standard time point is this if falling into the detection data record of the period
The average value of interior all detection data records;
(4) after carrying out time point amendment according to above-mentioned three principle, whether with the presence of 0 value may know that this group of number by data markers
According to the presence or absence of shortage of data point.
Step 2, data-privacyization processing: the letter in crowdsourcing track data, inevitably comprising individual subscriber mark
Breath, needs to carry out privacy processing to it.To the crowdsourcing track data collection for having been provided with authority data format and standard time point
In strong identity property, such as vehicle ID, user mobile phone condition code carry out anonymous processing, and it is original to generate one group of random digit substitution
Data;
Step 3, abnormal track data identification and processing: the data for exceeding parameter threshold are rejected, missing data is filled out
It studies for a second time courses one has flunked again, to form complete data set, the specific steps are as follows:
Step 3.1 carries out threshold decision, including threshold speed and vehicle running path threshold value, to the data for being more than threshold value
It is rejected;
(1) threshold speed
Since vehicle driving is on section, thus Vehicle Speed is by shadows such as road conditions, service level, control strategies
It ringing, the car speed uploaded in crowdsourcing track data be the instantaneous velocity at acquisition moment, there are overspeed situation, thus threshold speed
It is modified using correction factor f.It is final to determine the reasonable threshold range of speed are as follows:
0≤v≤fVmax
In formula: v indicates speed
F indicates correction factor, generally takes 1.3~1.5
VmaxIt indicates that road limits speed, is generally determined by road conditions, service level, control strategy etc.
(2) vehicle running path threshold value
Vehicle running path refers to longitude and latitude of the vehicle on section, due to communications link signals failure or positioning event
Barrier causes the vehicle geographical location of acquisition not in highway network limit, and actual conditions have deviation, thus needs to carry out geographical
Location matches.
Vehicle running path threshold value needs in advance to be acquired the longitude and latitude at highway network limit, forms road network geography position
Database is set, carries out appropriate widen using factor pair road network limit is widened.The allowed band of vehicle running path threshold value are as follows:
[x,y]∈{xlimit+ε,ylimit+ε}
In formula: [x, y] is the longitude and latitude of vehicle present position
The factor is widened in ε expression
{xlimit,ylimitIndicate highway network limit longitude and latitude data acquisition system
{xlimit+ε,ylimit+ ε } it indicates to widen rear highway network limit longitude and latitude data acquisition system
For crowdsourcing track data, if geographical location locating for vehicle is incited somebody to action not in the threshold range that driving path allows
The data markers are 0, judge it for data fault;If similarly data speed value is more than given threshold speed, by the data mark
It is denoted as 0, judges that it, for data fault, is rejected fault data.
Step 3.2 carries out missing data to fill up reparation, the time point missing data and step identified for step 1.3
Caused missing data is filled up after rapid 3.1 rejecting, and two ways is as follows:
(1) data filling is preferentially matched according to offline historical data base, i.e. the adjacent data in selection missing number strong point
Origin And Destination of the point as track, chooses identical Origin And Destination to obtain k track in offline historical data base,
The orbit segment of missing is determined according to interlude difference:
MinT=| Δ t- Δ t'i| i=(1,2 ..., k)
X (t)=x (ti)
In formula: Δ t indicates the time difference between missing data consecutive number strong point;
Δt'iIndicate the i-th track terminus time difference obtained from offline historical data base
(2) if historical data is insufficient, interpolation fitting is carried out using the data of adjacent time point before and after missing point, chooses phase
Each k known data point before and after adjacent time point, using Local Polynomial method construction multinomial, according to 2k known point calculate to
Coefficient value is determined, then calculate the corresponding position values of missing time points to restore track, to obtain ideal vehicle geographical location
Information.By the ideal data with its locating for the longitude and latitude data set of road network limit be compared, if the point is not belonging to limit data
Collection is then matched it to perpendicular to the position intersected at the adjacent two o'clock line of missing point with lane center:
In formula: aiFor undetermined coefficient.
Step 4, data set track pre-generatmg and adjustment: the complete crowdsourcing track data obtained after being handled by preceding 3 step
It is matched with road network, extracts same random number word (indicating identical vehicle ID), be ranked up according to time order and function, it will be through
Latitude is accurately matched with road network position, is connected adjacent data point and is formed entire run track.The track tentatively generated is adjusted again
It is whole, it rectified a deviation, denoised, ultimately generate complete smooth track, the specific steps are as follows:
Step 4.1 analyzes the track tentatively generated, for the data point obviously jumped, if before and after analyzing the point
The track that dry data are formed judges the point if the traffic direction of two sections of tracks is roughly the same for noise, rejects the point and utilizes
Interpolation method fills up data, corrects this section of track;If the traffic direction angle that the data point obviously jumped and consecutive number strong point are constituted
Greater than 90 °, and section multiple spot is constituted before and after the jump track traffic direction is opposite or angle is less than 90 °, then illustrates that the point is
Trajectory segment point, retains the data point;
Step 4.2 corrects part deviation data in track, such as the road limit in road limit and after widening
Between data point, match it to the position intersected perpendicular to vehicle traffic direction with section center line by this;
Step 4.3 carries out smooth trajectory processing using B-spline Curve;
Step 5, complete trajectory present and update historical data base.
It advantages of the present invention and has the active effect that
(1) a kind of traffic data quality control method based on crowdsourcing track data proposed by the present invention is adapted to vehicle rail
The processing of mark data compensates for existing method missing;
(2) present invention has carried out effective field extraction in advance in method of quality control, data redundancy is removed, for magnanimity
Original traffic data effectively save memory space, have unified the input specification of data, have kept data cleansing process easier;
(3) present invention handles the tiny time error as caused by data collection facility, when foring standard
Between the period, can effectively avoid the acquisition of its data from frequently and unevenly being difficult to be formed unification especially for crowdsourcing track data
The problem of standard time point;
(4) fault data present in traffic data of the present invention for crowdsourcing track data, missing data propose to understand
Certainly method, it is ensured that the integrality of track data provides accurate data supporting for traffic programme, traffic forecast.
Detailed description of the invention
Fig. 1 is the traffic data quality control method specific flow chart of the invention based on crowdsourcing track data;
Fig. 2 is path threshold zone of reasonableness schematic diagram provided in an embodiment of the present invention;
Fig. 3 is that missing orbit segment provided in an embodiment of the present invention matches schematic diagram;
Fig. 4 is that track data point provided in an embodiment of the present invention repairs schematic diagram.
Specific embodiment
Below in conjunction with drawings and examples, the present invention is described in further detail.
As shown in Figure 1, the embodiment of the present invention provides a kind of method of quality control of crowdsourcing track data, comprising:
Step 1, data cleansing: the collected data of device be will test and carry out key message extraction, recorded according to the format of agreement
Enter into database, detects acquired data attribute format, field contents, data set size, the steps include:
Step 1.1, assume crowdsourcing track data integrate field as G=VehicleID, OrderID, Time,
Longitude, Latitude }, in the data for carrying out not needing order ID when data analysis, which is rejected,
Designated time typing format are as follows: xxxx-xx-xxxx:xx:xx, longitude and latitude typing retain 6 decimals by agreement, if format is not
Match, then the data are labeled as 0, judges it for data fault;
Step 1.2 successively differentiates each critical field of data, detects whether there are certain field contents to be empty or lattice
Field contents are that messy code can not identify in the matched situation of formula, are 0 by the incomplete data markers of field contents, judge it for number
According to failure;
Step 1.3, the correction of standard time point and detection data collection size: by 24 hours one day according to equipment detection cycle T
It is divided into period standard time for time interval, sets the time interval of crowdsourcing track data upload as 1 minute, then standard time
Point be respectively 60s, 120s, 180s ..., allow acquisition time postpone be 30s, statistics gatherer data period be (30,90), (90,
150) ... interior data carry out time point amendment, using the data in the period as the data of standard time point 60s, 120s etc.,
Principle is as follows:
(1) if the detection data for falling into the period is recorded as 0, this standard time point data missing, by the point data mark
It is denoted as 0, judges it for shortage of data;
(2) if the detection data for falling into the period is recorded as 1, the numerical value of this standard time point is unique detection data
The numerical value of record:
D (t)=d (t)
(3) have n item (n > 1) if falling into the detection data record of the period, the numerical value of this standard time point is the period
The average value of all detection data records in range:
(4) after carrying out time point amendment according to above-mentioned three principle, whether with the presence of 0 value may know that this group of number by data markers
According to the presence or absence of shortage of data point.
Step 2, data-privacyization processing: to the crowdsourcing track data for having been provided with authority data format and standard time point
The strong identity property concentrated, such as vehicle ID, user mobile phone condition code carry out anonymous processing, and it is former to generate one group of random digit substitution
Beginning data;
Step 3, abnormal track data identification and processing: the data for exceeding parameter threshold are rejected, missing data is filled out
It studies for a second time courses one has flunked again, to form complete data set, the specific steps are as follows:
Step 3.1 carries out threshold decision, including threshold speed and vehicle running path threshold value, to the data for being more than threshold value
It is rejected;
(1) threshold speed
Speed is the instantaneous velocity and the real-time uploading speed of move vehicle when vehicle passes through section detector, and normal condition is all
It should be within the scope of road speed limit, but since section detector detection time is shorter thus there may be hypervelocity phenomenon, while many
The car speed uploaded in packet track data is the instantaneous velocity for acquiring the moment, and there is also overspeed situations, thus threshold speed is adopted
It is modified with correction factor f.It here is 80km/h by way of road speed limit value, correction factor is taken as 1.3, is obtained by correction factor
Speed is 104km/h.It is final to determine the reasonable threshold range of speed are as follows:
0≤v≤104km/h
(2) vehicle running path threshold value
As shown in Fig. 2, vehicle running path is by vehicle, the longitude and latitude of 4 points on section is constituted, and position 1 and position 4 are equal
In normal road network limit, due to communications link signals failure or positioning failure, lead to the vehicle geographical location 2,3 of acquisition
Not in highway network limit, and actual conditions have deviation, thus need to carry out geographical location matching.
Edge detection is carried out to travel, obtains longitude and latitude data at limit, forms road network limit geographic position data
Library carries out appropriate widen using factor pair road network limit is widened for slight error that may be present.Vehicle running path threshold value
Allowed band such as Fig. 2 widen shown in road network limit.
Vehicle geographical location 2 is observed, locating geographical location in the threshold range of road network limit permission, is incited somebody to action not after widening
The data markers are 0, judge that it, for data fault, and rejects the data;Observe vehicle geographical location 3, locating geographical location
In the threshold range allowed not in highway network limit but still in road network limit after widening, then it is 0 by the data markers, sentences
It break as data deviation.
Step 3.2 carries out missing data to fill up reparation, the time point missing data and step identified for step 1.3
Caused missing data is filled up after rapid 3.1 rejecting:
(1) data filling is preferentially matched according to offline historical data base, i.e. the adjacent data in selection missing number strong point
Origin And Destination of the point as track, chooses identical Origin And Destination to obtain k track in offline historical data base,
As shown in figure 4, according to 2 historical tracks that the radiation ring-type road network topology structure and missing number strong point adjacent data obtain,
Track 1 selects suburb loop wire to reach home, and the selection of track 2 is reached home from main city zone, determines missing according to interlude difference
Orbit segment, wherein the time difference for matching terminus in track 1 is 2min, the time difference for matching terminus in track 2 is 3min, and
The missing data consecutive points time difference is 2min, that is, is had
MinT=| Δ t- Δ ti' | i=(1,2)
That is matching track 1 and missing orbit segment is the most similar, therefore assigns the intermediate data value of track 1 to missing number strong point.
(2) if historical data is insufficient, interpolation fitting is carried out using the data of adjacent time point before and after missing point, chooses phase
Each k known data point before and after adjacent time point constructs multinomial using Local Polynomial method, so that it is geographical to obtain ideal vehicle
Location information, by the ideal data with its locating for the longitude and latitude data set of road network limit be compared, if the point is not belonging to limit
Data set is then matched it to perpendicular to the position intersected at the adjacent two o'clock line of missing point with lane center:
In formula: aiFor undetermined coefficient.Coefficient value is calculated according to 2k known point, then calculates the corresponding positions of missing time points
Track can be restored by setting value.
Step 4, data set track pre-generatmg and adjustment: the complete crowdsourcing track data obtained after being handled by preceding 3 step
It is matched with road network, extracts same random number word (indicating identical vehicle ID), be ranked up according to time order and function, it will be through
Latitude is accurately matched with road network position, is connected adjacent data point and is formed entire run track.The track tentatively generated is adjusted again
It is whole, it rectified a deviation, denoised, ultimately generate complete track, the specific steps are as follows:
Step 4.1 analyzes the track tentatively generated, for the data point obviously jumped, if before and after analyzing the point
The track that dry data are formed judges the point if the traffic direction of two sections of tracks is roughly the same for noise, rejects the point and utilizes
Interpolation method fills up data, corrects this section of track;If the folder that two sections of track traffic directions before and after the data point obviously jumped are formed
Angle is greater than 90 °, then illustrates that the point is trajectory segment point, retain the data point;
Step 4.2 corrects part deviation data in track, such as the road limit in road limit and after widening
Between data point, match it to the position intersected perpendicular to vehicle traffic direction with section center line by this;
Step 4.3 carries out smooth trajectory processing using B-spline Curve;
Step 5, complete trajectory present and update historical data base.
Claims (3)
1. a kind of method of quality control of " crowdsourcing " track data, which is characterized in that the described method includes:
Step 1: data cleansing step
" crowdsourcing " track data is subjected to key message extraction, is entered into database according to the format of agreement, data mark is carried out
Quasi- time point corrects and detects acquired data attribute format, field contents, data set size;
Step 2: abnormal track data identification and processing step
The data for exceeding parameter threshold are rejected, missing data are carried out to fill up reparation, to form complete data set, comprising:
S201 threshold decision sub-step carries out threshold speed and vehicle running path threshold decision, to be more than threshold value data into
Row is rejected;
Threshold speed is modified using correction factor f, it is final to determine the reasonable threshold range of speed are as follows: 0≤v≤fVmax,
In formula: v indicates speed, and f indicates correction factor, VmaxIndicate that road limits speed;If " crowdsourcing " track data velocity amplitude is more than to give
The data markers are 0, judge that it, for data fault, is rejected fault data by fixed threshold speed;For vehicle driving
Path threshold is widened using factor pair road network limit progress appropriateness is widened, the allowed band of vehicle running path threshold value are as follows: [x, y]
∈{xlimit+ε,ylimit+ ε } in formula [x, y] be the longitude and latitude of vehicle present position, the factor is widened in ε expression, { xlimit,ylimitTable
Show highway network limit longitude and latitude data acquisition system, { xlimit+ε,ylimit+ ε } it indicates to widen rear highway network limit longitude and latitude data
Set;For " crowdsourcing " track data, if geographical location locating for vehicle not in the threshold range that driving path allows, should
Data markers are 0, judge it for data fault;
S202 carries out missing data to fill up reparation, for the data of time point missing in " crowdsourcing " track data and passes through threshold value
Caused missing data is filled up after judgement processing is rejected, and two ways is as follows:
Matched first according to offline historical data base, choose the adjacent data point in missing number strong point as the starting point of track and
Terminal is chosen identical Origin And Destination in offline historical data base to obtain k track, is determined according to interlude difference
The orbit segment of missing: minT=| Δ t- Δ t'i| i=(1,2 ..., k);X (t)=x (ti) in formula: Δ t indicates missing data phase
Time difference between adjacent data point;Δti' indicate the i-th track terminus time difference obtained from offline historical data base;
If historical data is insufficient, interpolation fitting is carried out using the data of adjacent time point before and after missing point, before choosing adjacent time point
Each k known data point afterwards constructs multinomial using Local Polynomial method, calculates undetermined coefficient value according to 2k known point,
The corresponding position value for calculating missing time points again can restore track, so that ideal vehicle geographical location information is obtained, by this
Ideal data is compared with the longitude and latitude data set of its locating road network limit, by it if the point is not belonging to limit data set
It is fitted on perpendicular to the position intersected at the adjacent two o'clock line of missing point with lane centerIn formula: i=0,1,
2 ..., 2k-1, aiFor undetermined coefficient;The time at t expression missing number strong point;tiIndicate choose i-th of known data point when
Between.
Step 3: data set track pre-generatmg and set-up procedure
Complete " crowdsourcing " track data obtained after processing is matched with road network;Extracting indicates that identical vehicle ID's is identical
Random digit is ranked up according to time order and function, and longitude and latitude is accurately matched with road network position, adjacent data point is connected and is formed
Entire run track;The track tentatively generated is adjusted again, rectified a deviation, denoised, ultimately generates complete smooth track, specifically
Steps are as follows:
S301 analyzes the track tentatively generated, for the data point of jump, analyzes what several data in point front and back were formed
Track, if the traffic direction of the two sections of tracks in front and back is consistent, which is judged as noise, rejects the point and is filled up using interpolation method
Data correct this section of track;If the traffic direction angle that the data point of jump and consecutive number strong point are constituted is greater than 90 °, and the jump
The track traffic direction that section multiple spot is constituted before and after hop is opposite or angle is less than 90 °, then illustrates that the point is trajectory segment point, retain
The data point;
S302 corrects part deviation data in track, such as the number between the road limit in road limit and after widening
Strong point matches it to the position intersected perpendicular to vehicle traffic direction with section center line by this;
S303 carries out smooth trajectory to each estimated data point using B-spline Curve and handles to obtain new " crowdsourcing " track;
Step 4: complete trajectory presents and updates historical data base.
2. the method for quality control of one kind " crowdsourcing " track data according to claim 1, which is characterized in that in the step
Between rapid one and step 2 further include: data-privacy processing step, to having been provided with authority data format and standard time point
The private data concentrated of crowdsourcing track data, generate one group of random digit and substitute initial data.
3. the method for quality control of one kind " crowdsourcing " track data according to claim 1, which is characterized in that in the step
Rapid one includes:
The data that S101 will acquire carry out key feature extraction, reject redundant field, and set data inputting format according to agreement,
The attribute integrality of inspection data;The standard typing format that each field is set before data inputting database, if format mismatching,
The data are then labeled as 0, judge it for data fault;
S102 successively differentiates each critical field of data, detects whether there are certain field contents to be empty or format match
In the case of field contents be messy code can not identify, by the incomplete data markers of field contents be 0, judge it for data fault;
The correction of S103 standard time point and detection data collection size;By 24 hours one day according to equipment detection cycle T between the time
Every being divided into period standard time, established standards time point is t', and preset equipment detection cycle is T, when allowing to acquire
Between delay be t, statistics gatherer data period be (t'+t-T, t'+t) in data carry out time point amendment;If falling into the period
Detection data be recorded as 0, then this standard time point data lack, by the point data be labeled as 0, judge it for shortage of data;
If the detection data for falling into the period is recorded as 1, the numerical value of this standard time point is the numerical value of unique detection data record;
If the detection data record for falling into the period has a plurality of, all testing numbers in segment limit when the numerical value of this standard time point is this
According to the average value of record;
Carry out time point amendment after, by data markers whether with the presence of 0 value can determine this group of data whether there is shortage of data point.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811227001.4A CN109215347B (en) | 2018-10-22 | 2018-10-22 | Traffic data quality control method based on crowdsourcing trajectory data |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811227001.4A CN109215347B (en) | 2018-10-22 | 2018-10-22 | Traffic data quality control method based on crowdsourcing trajectory data |
Publications (2)
Publication Number | Publication Date |
---|---|
CN109215347A true CN109215347A (en) | 2019-01-15 |
CN109215347B CN109215347B (en) | 2020-10-09 |
Family
ID=64981211
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201811227001.4A Expired - Fee Related CN109215347B (en) | 2018-10-22 | 2018-10-22 | Traffic data quality control method based on crowdsourcing trajectory data |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109215347B (en) |
Cited By (24)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110008196A (en) * | 2019-04-17 | 2019-07-12 | 华南理工大学 | A kind of method for sorting and device, equipment and medium of magnanimity traffic point data |
CN110222131A (en) * | 2019-05-21 | 2019-09-10 | 北京交通大学 | The beginning and the end information extracting method and device |
CN110647533A (en) * | 2019-09-17 | 2020-01-03 | 同恩(上海)工程技术有限公司 | Data monitoring method, system and storage medium for structure monitoring system |
CN110704409A (en) * | 2019-09-21 | 2020-01-17 | 广东毓秀科技有限公司 | Method for optimizing rail transit data quality through data triangular prism algorithm |
CN110781266A (en) * | 2019-09-16 | 2020-02-11 | 北京航空航天大学 | Urban perception data processing method based on time-space causal relationship |
CN110986970A (en) * | 2019-12-21 | 2020-04-10 | 武汉中海庭数据技术有限公司 | Method and system for optimizing positions of track points and traffic signs |
CN111008672A (en) * | 2019-12-23 | 2020-04-14 | 腾讯科技(深圳)有限公司 | Sample extraction method, sample extraction device, computer-readable storage medium and computer equipment |
CN111192452A (en) * | 2019-12-27 | 2020-05-22 | 东软集团股份有限公司 | Stroke data segmentation method and device, storage medium and electronic equipment |
CN111189459A (en) * | 2020-01-10 | 2020-05-22 | 成都信息工程大学 | Method and device for matching positioning information with road |
CN111198562A (en) * | 2019-12-31 | 2020-05-26 | 武汉中海庭数据技术有限公司 | Preprocessing optimization method for space line characteristics of crowdsourcing fragment map |
CN111288999A (en) * | 2020-02-19 | 2020-06-16 | 深圳大学 | Pedestrian road network attribute detection method, device and equipment based on mobile terminal |
CN111653115A (en) * | 2019-03-04 | 2020-09-11 | 腾讯大地通途(北京)科技有限公司 | Task dividing method and device and storage medium |
CN111833616A (en) * | 2020-07-10 | 2020-10-27 | 北京交通大学 | Data quality control method and device for overload recognition of internet-connected truck |
CN112013862A (en) * | 2020-07-31 | 2020-12-01 | 深圳大学 | Pedestrian network extraction and updating method based on crowdsourcing trajectory |
CN112148811A (en) * | 2019-06-26 | 2020-12-29 | 陕西汽车集团有限责任公司 | Vehicle-mounted GPS track path compression method |
CN112382090A (en) * | 2020-11-11 | 2021-02-19 | 北京百度网讯科技有限公司 | Method, apparatus, device and storage medium for outputting information |
CN112380065A (en) * | 2020-11-18 | 2021-02-19 | 北京百度网讯科技有限公司 | Data restoration method and device, electronic equipment and storage medium |
CN112527943A (en) * | 2020-12-25 | 2021-03-19 | 北京百度网讯科技有限公司 | Map rendering method, map rendering device, map rendering equipment and storage medium |
CN112748452A (en) * | 2020-12-11 | 2021-05-04 | 上海城市交通设计院有限公司 | GPS track cleaning method based on road network data |
CN112816954A (en) * | 2021-02-09 | 2021-05-18 | 中国信息通信研究院 | Road side perception system evaluation method and system based on truth value |
CN113553390A (en) * | 2021-07-12 | 2021-10-26 | 北京清环宜境技术有限公司 | Running GIS track deviation rectifying method for vehicle-mounted atmospheric micro station |
CN114093164A (en) * | 2021-11-15 | 2022-02-25 | 上海市城乡建设和交通发展研究院 | Dynamic traffic flow identification and correction method, device and equipment based on vehicle track |
CN114743406A (en) * | 2022-03-11 | 2022-07-12 | 中国电子科技集团公司第五十四研究所 | Ship track entanglement removal method |
CN115657647A (en) * | 2022-12-19 | 2023-01-31 | 小米汽车科技有限公司 | Fault determination method, device, vehicle and storage medium |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101373559A (en) * | 2007-08-24 | 2009-02-25 | 同济大学 | Method for evaluating city road net traffic state based on floating vehicle data |
CN104574967A (en) * | 2015-01-14 | 2015-04-29 | 合肥革绿信息科技有限公司 | City large-area road network traffic sensing method based on plough satellite |
CN107845259A (en) * | 2017-10-24 | 2018-03-27 | 东南大学 | Public transport operation situation real-time feedback system and public transport real-time running data processing method |
CN108492562A (en) * | 2018-04-12 | 2018-09-04 | 连云港杰瑞电子有限公司 | Intersection vehicles trajectory reconstruction method based on fixed point detection with the alert data fusion of electricity |
CN108665704A (en) * | 2018-05-17 | 2018-10-16 | 北京航空航天大学 | Intersection period maximum queue length method of estimation based on crowdsourcing track data |
-
2018
- 2018-10-22 CN CN201811227001.4A patent/CN109215347B/en not_active Expired - Fee Related
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101373559A (en) * | 2007-08-24 | 2009-02-25 | 同济大学 | Method for evaluating city road net traffic state based on floating vehicle data |
CN104574967A (en) * | 2015-01-14 | 2015-04-29 | 合肥革绿信息科技有限公司 | City large-area road network traffic sensing method based on plough satellite |
CN107845259A (en) * | 2017-10-24 | 2018-03-27 | 东南大学 | Public transport operation situation real-time feedback system and public transport real-time running data processing method |
CN108492562A (en) * | 2018-04-12 | 2018-09-04 | 连云港杰瑞电子有限公司 | Intersection vehicles trajectory reconstruction method based on fixed point detection with the alert data fusion of electricity |
CN108665704A (en) * | 2018-05-17 | 2018-10-16 | 北京航空航天大学 | Intersection period maximum queue length method of estimation based on crowdsourcing track data |
Cited By (36)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111653115A (en) * | 2019-03-04 | 2020-09-11 | 腾讯大地通途(北京)科技有限公司 | Task dividing method and device and storage medium |
CN111653115B (en) * | 2019-03-04 | 2022-02-15 | 腾讯大地通途(北京)科技有限公司 | Task dividing method and device and storage medium |
CN110008196A (en) * | 2019-04-17 | 2019-07-12 | 华南理工大学 | A kind of method for sorting and device, equipment and medium of magnanimity traffic point data |
CN110222131A (en) * | 2019-05-21 | 2019-09-10 | 北京交通大学 | The beginning and the end information extracting method and device |
CN112148811B (en) * | 2019-06-26 | 2023-01-10 | 陕西汽车集团股份有限公司 | Vehicle-mounted GPS track path compression method |
CN112148811A (en) * | 2019-06-26 | 2020-12-29 | 陕西汽车集团有限责任公司 | Vehicle-mounted GPS track path compression method |
CN110781266B (en) * | 2019-09-16 | 2020-06-09 | 北京航空航天大学 | Urban perception data processing method based on time-space causal relationship |
CN110781266A (en) * | 2019-09-16 | 2020-02-11 | 北京航空航天大学 | Urban perception data processing method based on time-space causal relationship |
CN110647533A (en) * | 2019-09-17 | 2020-01-03 | 同恩(上海)工程技术有限公司 | Data monitoring method, system and storage medium for structure monitoring system |
CN110704409A (en) * | 2019-09-21 | 2020-01-17 | 广东毓秀科技有限公司 | Method for optimizing rail transit data quality through data triangular prism algorithm |
CN110986970A (en) * | 2019-12-21 | 2020-04-10 | 武汉中海庭数据技术有限公司 | Method and system for optimizing positions of track points and traffic signs |
CN111008672A (en) * | 2019-12-23 | 2020-04-14 | 腾讯科技(深圳)有限公司 | Sample extraction method, sample extraction device, computer-readable storage medium and computer equipment |
CN111192452B (en) * | 2019-12-27 | 2021-06-18 | 东软集团股份有限公司 | Stroke data segmentation method and device, storage medium and electronic equipment |
CN111192452A (en) * | 2019-12-27 | 2020-05-22 | 东软集团股份有限公司 | Stroke data segmentation method and device, storage medium and electronic equipment |
CN111198562A (en) * | 2019-12-31 | 2020-05-26 | 武汉中海庭数据技术有限公司 | Preprocessing optimization method for space line characteristics of crowdsourcing fragment map |
CN111189459B (en) * | 2020-01-10 | 2023-12-22 | 成都信息工程大学 | Method and device for matching positioning information with road |
CN111189459A (en) * | 2020-01-10 | 2020-05-22 | 成都信息工程大学 | Method and device for matching positioning information with road |
CN111288999B (en) * | 2020-02-19 | 2021-08-31 | 深圳大学 | Pedestrian road network attribute detection method, device and equipment based on mobile terminal |
CN111288999A (en) * | 2020-02-19 | 2020-06-16 | 深圳大学 | Pedestrian road network attribute detection method, device and equipment based on mobile terminal |
CN111833616A (en) * | 2020-07-10 | 2020-10-27 | 北京交通大学 | Data quality control method and device for overload recognition of internet-connected truck |
CN111833616B (en) * | 2020-07-10 | 2021-11-16 | 北京交通大学 | Data quality control method and device for overload recognition of internet-connected truck |
CN112013862A (en) * | 2020-07-31 | 2020-12-01 | 深圳大学 | Pedestrian network extraction and updating method based on crowdsourcing trajectory |
CN112382090A (en) * | 2020-11-11 | 2021-02-19 | 北京百度网讯科技有限公司 | Method, apparatus, device and storage medium for outputting information |
CN112380065A (en) * | 2020-11-18 | 2021-02-19 | 北京百度网讯科技有限公司 | Data restoration method and device, electronic equipment and storage medium |
CN112380065B (en) * | 2020-11-18 | 2023-08-29 | 北京百度网讯科技有限公司 | Data restoration method and device, electronic equipment and storage medium |
CN112748452B (en) * | 2020-12-11 | 2022-09-23 | 上海城市交通设计院有限公司 | GPS track cleaning method based on road network data |
CN112748452A (en) * | 2020-12-11 | 2021-05-04 | 上海城市交通设计院有限公司 | GPS track cleaning method based on road network data |
CN112527943A (en) * | 2020-12-25 | 2021-03-19 | 北京百度网讯科技有限公司 | Map rendering method, map rendering device, map rendering equipment and storage medium |
CN112527943B (en) * | 2020-12-25 | 2023-12-01 | 北京百度网讯科技有限公司 | Map rendering method, device, equipment and storage medium |
CN112816954A (en) * | 2021-02-09 | 2021-05-18 | 中国信息通信研究院 | Road side perception system evaluation method and system based on truth value |
CN112816954B (en) * | 2021-02-09 | 2024-03-26 | 中国信息通信研究院 | Road side perception system evaluation method and system based on true value |
CN113553390A (en) * | 2021-07-12 | 2021-10-26 | 北京清环宜境技术有限公司 | Running GIS track deviation rectifying method for vehicle-mounted atmospheric micro station |
CN114093164B (en) * | 2021-11-15 | 2022-08-19 | 上海市城乡建设和交通发展研究院 | Dynamic traffic flow identification and correction method, device and equipment based on vehicle track |
CN114093164A (en) * | 2021-11-15 | 2022-02-25 | 上海市城乡建设和交通发展研究院 | Dynamic traffic flow identification and correction method, device and equipment based on vehicle track |
CN114743406A (en) * | 2022-03-11 | 2022-07-12 | 中国电子科技集团公司第五十四研究所 | Ship track entanglement removal method |
CN115657647A (en) * | 2022-12-19 | 2023-01-31 | 小米汽车科技有限公司 | Fault determination method, device, vehicle and storage medium |
Also Published As
Publication number | Publication date |
---|---|
CN109215347B (en) | 2020-10-09 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109215347A (en) | A kind of traffic data quality control method based on crowdsourcing track data | |
CN110008872B (en) | Road network extraction method combining vehicle track and remote sensing image | |
CN109584553B (en) | Road section relevance missing completion method based on space-time information | |
CN111462484B (en) | Congestion state determination method, device, equipment and computer readable storage medium | |
CN104813143B (en) | Road network analyzer, road network analysis system and road network analysis method | |
CN101719315B (en) | Method for acquiring dynamic traffic information based on middleware | |
CN103646187B (en) | Method for obtaining vehicle travel path and OD (Origin-Destination) matrix in statistic period | |
CN102394008B (en) | Assessing road traffic conditions using data from mobile data sources | |
CN109688532B (en) | Method and device for dividing city functional area | |
CN109754594A (en) | A kind of road condition information acquisition method and its equipment, storage medium, terminal | |
CN106488405B (en) | A kind of position predicting method of fusion individual and neighbour's movement law | |
CN107563566B (en) | Inter-bus-station operation time interval prediction method based on support vector machine | |
CN108171993B (en) | Highway vehicle speed calculation method based on mobile phone signaling big data | |
CN105632222B (en) | Forecast the method and its system of arrival time | |
CN104884899A (en) | Method of determining trajectories through one or more junctions of a transportation network | |
US20060111833A1 (en) | Method and system for modeling and processing vehicular traffic data and information and applying thereof | |
CN108362293A (en) | A kind of track of vehicle matching process based on key point technology | |
CN107330085B (en) | Method for judging, identifying and correcting error position of fixed sensor in big data environment | |
CN103942312B (en) | Bus transfer lines planning method and device | |
CN109859495A (en) | A method of overall travel speed is obtained based on RFID data | |
CN108090722A (en) | A kind of logistics network station recognition equipment and recognition methods based on GPS | |
CN102622879A (en) | Traffic information providing apparatus | |
CN106997666A (en) | A kind of method that utilization mobile phone signaling data position switching obtains traffic flow speed | |
CN104422451A (en) | Road recognition method and road recognition apparatus | |
CN105096590A (en) | Traffic information generation method and device |
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: 20201009 Termination date: 20211022 |