Summary of the invention
It is an object of the present invention to overcome the above-mentioned drawbacks of the prior art and provide one kind to be based on floating track
The compound road network bottleneck point recognition methods of mark data.
The purpose of the present invention can be achieved through the following technical solutions:
A kind of compound road network bottleneck point recognition methods based on Floating Car track data, comprising:
Step S1: data cleansing is carried out to Floating Car track data;
Step S2: merging for track data and geographic information data is carried out to the Floating Car track data after over cleaning;
Step S3: it carries out information excavating and obtains bottleneck point;
Step S4: bottleneck point is divided by moving bottleneck or fixed bottleneck according to the change in location of bottleneck point, and according to bottle
Bottleneck point is divided into often hair property bottleneck or sporadic bottleneck by the time change of neck point;
Step S5: the output bottleneck point origin cause of formation.
The step S1 is specifically included:
Step S11: judging whether the vehicle ID of current trace points in Floating Car track data, timestamp lack or abnormal,
If it is, S12 is thened follow the steps, conversely, thening follow the steps S13;
Step S12: current trace points are abandoned, next tracing point is read and judges whether track data terminates, if it has not, then
Execute step S11;
Step S13: judging whether the instantaneous velocity of current trace points lacks or abnormal, if it is, S14 is thened follow the steps,
Conversely, thening follow the steps S15;
Step S14: being attempted smoothly to be repaired using the instantaneous velocity of adjacent track point, and judge whether success, if it has, then
Step S15 is executed, conversely, thening follow the steps S12;
Step S15: judge whether the longitude and latitude of current trace points lacks or abnormal, if it is, thening follow the steps S16, instead
It, thens follow the steps S17;
Step S16: it attempts smoothly to repair using the longitude and latitude of adjacent track point, and judges whether success, if it has, then holding
Row step S17, conversely, thening follow the steps S12;
Step S17: reading next tracing point and judges whether track data terminates, if it has not, thening follow the steps S11.
Judge whether successful process in the step S14 specifically: judging the difference of the instantaneous velocity of adjacent track point is
It is no to be less than threshold velocity.
Judge whether successful process in the step S16 specifically:
The distance between two tracing points of calculation of longitude & latitude based on adjacent track point;
First Speed is calculated according to the distance between two tracing points and time difference, and judges the First Speed and wink
Whether the difference of Shi Sudu is less than setting ratio, if, then it is assumed that success, it is on the contrary then think failure.
The step S2 is specifically included:
Step S21: by delimiting a rectangular area, tracing point all in target intersection group is obtained;
Step S22: it assigns tracing point intersection number interid, road number roadid, go out row number tripid;
Step S23: direction of travel when whether passing through intersection according to track and passing through intersection assigns tracing point and turns
To number moveid;
Step S24: for the section between two intersections, tracing point is being repaired just according to the intersection that vehicle is driven towards
Intersection number interid;
Step S25: the distance of vehicle distances stop line is obtained.
The step S22 is specifically included:
Step S221: the second is converted by road time format;
Step S222: for the road between intersection, centre adheres to two intersections separately to half-open, assigns and handing over for each tracing point
Prong number interid;
Step S223: delimiting the region of longitude and latitude for different sections of highway, and correspondence invests road number roadid value;
Step S224: row number tripid is assigned out according to the corresponding row order out of each tracing point.
It is specific to wrap for the tracing point that each tripid of each each car of each car is covered in the step S25
It includes:
If only including 1 roadid value, the distance for arriving intersection is directly calculated;
If containing 3 roadid values, an intersection only have passed through, equally directly calculate distance;
If characterization have passed through multiple intersections comprising being greater than 3 roadid values, a pair of cross mouth number is successively screened
The corresponding tracing point of interid, calculates separately the distance apart from stop line.
The step S5 includes:
If moving bottleneck, there is the vehicle (may be truck or car of going slowly) of retardance traffic in output;
If fixed bottleneck, the traffic capacity for exporting the existing means of transportation of the bottleneck point of the bottleneck point be can no longer meet
Status transport need.
Compared with prior art, the invention has the following advantages:
1) acquisition of track data need not additionally install Vehicle Detection sensor, and procurement cost is cheap.
2) the various topological structure and traffic flow condition for combining compound road network, classify to bottleneck point.The bottleneck of foundation
Feature judgement and the opposite more easily execution of classifying method.
Specific embodiment
The present invention is described in detail with specific embodiment below in conjunction with the accompanying drawings.The present embodiment is with technical solution of the present invention
Premised on implemented, the detailed implementation method and specific operation process are given, but protection scope of the present invention is not limited to
Following embodiments.
The purpose of the application is in order to overcome the shortcomings of existing bottleneck point identifying and analyzing method, based on the traffic to become increasingly abundant
Track big data constructs compound road network bottleneck point recognizer, to promote traffic condition detection accuracy, expand bottleneck point knowledge
Other application range.The it is proposed of this method peomotes urban traffic control system construction and upgrading, accurately identifies and deposits city
Traffic congestion critical issue, to reach the purpose for more targetedly alleviating urban traffic blocking.
It proposes compound road network bottleneck point recognition methods, is based on Floating Car track data, commented from the quantization of time-domain space angle
Estimate the parameter attribute and rule of bottleneck point, thus for network traffic supersaturation control evaluation provide with optimization it is objective, effective
Foundation.
A kind of compound road network bottleneck point recognition methods based on Floating Car track data, as shown in Figure 1, comprising:
Step S1: carrying out data cleansing to Floating Car track data, as shown in Fig. 2, specifically including:
Step S11: judging whether the vehicle ID of current trace points in Floating Car track data, timestamp lack or abnormal,
If it is, S12 is thened follow the steps, conversely, thening follow the steps S13;
Step S12: current trace points are abandoned, next tracing point is read and judges whether track data terminates, if it has not, then
Execute step S11;
Step S13: judging whether the instantaneous velocity of current trace points lacks or abnormal, if it is, S14 is thened follow the steps,
Conversely, thening follow the steps S15;
Step S14: being attempted smoothly to be repaired using the instantaneous velocity of adjacent track point, and judge whether success, if it has, then
Step S15 is executed, conversely, thening follow the steps S12;
Specifically, should judge whether that successful process was specifically as follows in the process: judging the instantaneous velocity of adjacent track point
Difference whether be less than threshold velocity.
Step S15: judge whether the longitude and latitude of current trace points lacks or abnormal, if it is, thening follow the steps S16, instead
It, thens follow the steps S17;
Step S16: it attempts smoothly to repair using the longitude and latitude of adjacent track point, and judges whether success, if it has, then holding
Row step S17, conversely, thening follow the steps S12;
Specifically, wherein judging whether successful process can include:
The distance between two tracing points of calculation of longitude & latitude based on adjacent track point;Then according between two tracing points
Distance and the time difference First Speed is calculated, and judge whether the difference of the First Speed and instantaneous velocity is less than setting ratio
Example, if, then it is assumed that success, it is on the contrary then think failure.
Step S17: reading next tracing point and judges whether track data terminates, if it has not, thening follow the steps S11.
Step S2: carrying out merging for track data and geographic information data to the Floating Car track data after over cleaning,
It specifically includes:
Step S21:: there is a large amount of unrelated tracing point in screening target area in file, by delimiting a rectangular area,
Obtain tracing point all in target intersection group;
Step S22: tracing point intersection number interid, road number roadid are assigned, goes out row number tripid, tool
Body includes:
Step S221: the second is converted by road time format, such as be converted into number of seconds in 35 seconds 35 minutes at 17 points;
Step S222: for the road between intersection, centre adheres to two intersections separately to half-open, assigns and handing over for each tracing point
Prong number interid;
Step S223: delimiting the region of longitude and latitude for different sections of highway, and correspondence invests road number roadid value, specifically,
The region of longitude and latitude delimited to different sections of highway, is corresponded and is assigned roadid value (intersection region can random assignment), then
Roadid is tracing point of the empty data line i.e. not on section, is deleted;
Step S224: row number tripid is assigned out according to the corresponding row order out of each tracing point, specifically, for one
Vehicle, it may have multiple trip in one day, and not occur continuously in time, and therefore, different trips is drawn
It separates.If time interval is greater than 15 seconds (official's explanation, the tracing point sample frequencys of type=11 between upper and lower two tracing points
It is 15 seconds/time, remaining is 1 second/time), then judge that the vehicle has started another new trip, tripid is numbered from 1, this is
Place mat has been made in the setting of subsequent turn around parameters.
Step S23: direction of travel when whether passing through intersection according to track and passing through intersection assigns tracing point and turns
To number moveid, specifically, as follows:
A) for having passed through the track of vehicle of intersection, setting: turning left is 1, and keeping straight on is 2, and turning right is 3;
B) for not passed through the track of vehicle of intersection, 4 are set as.
Step S24: for the section between two intersections, tracing point is being repaired just according to the intersection that vehicle is driven towards
Intersection number interid just needs to correct for the section between two intersections: which the intersection that vehicle drives towards is
As soon as, its interid is assigned a value of to the number of which intersection;
Step S25: the distance of vehicle distances stop line is obtained, wherein for each of each each car of each car
The tracing point that tripid is covered, specifically includes:
If only including 1 roadid value, the distance for arriving intersection is directly calculated;
If containing 3 roadid values, an intersection only have passed through, equally directly calculate distance;
If characterization have passed through multiple intersections comprising being greater than 3 roadid values, a pair of cross mouth number is successively screened
The corresponding tracing point of interid, calculates separately the distance apart from stop line.
Next, draw space-time diagram as shown in figure 3, check virtual stop line at a distance from true stop line, to it is existing away from
It is modified from data.
Step S3: it carries out information excavating and obtains bottleneck point;
Step S4: bottleneck point is divided by moving bottleneck or fixed bottleneck according to the change in location of bottleneck point, and according to bottle
Bottleneck point is divided into often hair property bottleneck or sporadic bottleneck by the time change of neck point;
Bottleneck point analysis specifically includes: bottleneck point Evolution, the analysis of bottleneck range of point influence.
(1) bottleneck point Evolution
Section is drawn and is divided into several sections of separate spaces at certain intervals, it is assumed that each unit road section traffic volume situation is identical, is based on
It corrects track data and carries out state matching.It detects that upstream speed is significantly lower than the section in downstream, is then bottleneck point.Record is different
When the changing rule of bottleneck point inscribed, can use machine learning algorithm and carry out feature identification and predicted with evolvement trend.
(2) bottleneck range of point influence
For the bottleneck point identified, its upstream traffic behavior is observed based on track data.Generally according to 1km, 5km range
The segmentation of carry out state (finds the initial point position of bottleneck upstream congestion), and state segmentation range is flexibly adjusted according to the range in city
It is whole.Since track positioning accuracy is not up to lane grade, need to judge by the deflection attribute of track herein.
Bottleneck type includes:
1) normal hair property bottleneck: more day datas repeat to recognize certain section, and at the close moment, there are bottleneck or bottleneck points
The coverage of mobile trend and bottleneck point is consistent, then can all be classified as the often property sent out bottleneck;
2) sporadic bottleneck: the bottleneck point and its coverage detected, spatial homing phase of getting along well in the historical data
Matching, temporal regularity of also getting along well match;
3) moving bottleneck and fixed bottleneck: being directed to certain bottleneck, if bottleneck point is not moved during being formed with dissipation
Dynamic, then otherwise it is moving bottleneck that explanation, which is fixed bottleneck,.
Step S5: the output bottleneck point origin cause of formation, comprising:
If moving bottleneck, there is the vehicle (may be truck or car of going slowly) of retardance traffic in output;
If fixed bottleneck, the traffic capacity for exporting the existing means of transportation of the bottleneck point of the bottleneck point be can no longer meet
Status transport need.