CN103325245A - Method for predicting space-time traveling track of blacklisted vehicle - Google Patents

Abstract

The invention relates to a method for predicting the space-time traveling track of a blacklisted vehicle. The method comprises the prediction of the traveling route of the blacklisted vehicle and the prediction of the traveling time of the blacklisted vehicle on the predicted traveling route. The prediction of the traveling route is achieved with a traveling route predicting method based on historical route similarity. The prediction of the traveling time of the blacklisted vehicle on the predicted traveling route is achieved with a traveling time predicting method based on road traffic condition evaluation. The method for predicting the space-time traveling track of the blacklisted vehicle is high in efficiency. According to the method for predicting the space-time traveling track of a blacklisted vehicle, the traveling route of the blacklisted vehicle can be predicted in real time after the blacklisted vehicle appears, the next position where the blacklisted vehicle might appear and the time when the blacklisted vehicle appears at the position can be determined, video surveillance and control can be conducted on the corresponding position in advance according to the predicted traveling route and the predicted traveling time of the blacklisted vehicle, and decision aids are provided for the arrest of the blacklisted vehicle.

Description

A kind of space-time driving trace Forecasting Methodology of blacklist vehicle

Technical field

The present invention relates to a kind of track of vehicle Forecasting Methodology, relate in particular to a kind of space-time driving trace Forecasting Methodology of blacklist vehicle.

Background technology

Along with the fast development of IT industry, advanced location, detection, monitoring and the communication technology are widely used in field of traffic.At present the highway deploy between each incity, city road bayonet socket and the city vehicle snapshot camera of magnanimity, can produce a large amount of Traffic Informations and vehicle pass-through information, recorded the driving trace of all vehicles; Smart mobile phone, panel computer etc. are with hand-held mobile terminal or the on-board driving recording unit of positioning function in addition, and each has been recorded by the collection vehicle driving trace in Real-time Collection GPS position.Analyze track of vehicle information and be applied in the focus of a research of intelligent transportation field, mainly concentrate on track of vehicle prediction, road section traffic volume status analysis, the analysis of road security of operation etc.

It is two-layer that vehicle space-time driving trace research is divided into, and one deck is the path space trajectory predictions of travelling future for vehicle, predicts namely which bar road vehicle can select away, by which traffic marking, and bayonet socket position for example; Another layer is the time locus prediction for following driving path, namely how long spends at the road of selecting, and estimates constantly to pass through at which the marker location of prediction.

The driving trace forecasting research at present a large amount of unknown paths all is based on GIS map and GPS navigation, according to the geography information of road and general destination choice shortest path, some algorithms are arranged in addition with reference to current congestion in road situation, select optimal path for running time is the shortest in current and future transportation situation.These methods were not considered vehicle driver's personal habits, for example the historical track information of vehicle.

The time locus prediction substantially all is at present the driving trace prediction for fixed route, and the arrival time prediction of the public transit vehicle of route is for example arranged; Also have in addition specially and predict for the driving trace in unknown path, such as taxi, common public vehicles etc., the starting point of known vehicle and destination, do not consider what path vehicle can select, just expend time in to predict how long arrive the destination needs according to a large amount of history runs, such as the driving trace of the people such as Balan R according to a large amount of taxis of Singapore, proposition calculate needed time and the expense of calling a taxi based on similar historical journey time and expense.

Summary of the invention

Technical matters to be solved by this invention is the deficiency for existing Gate System, a kind of space-time driving trace Forecasting Methodology of blacklist vehicle is provided, the method can not only be predicted the driving path of blacklist vehicle, can also predict the running time of this blacklist vehicle on the driving path of prediction.

The technical scheme that the present invention solves the problems of the technologies described above is as follows: a kind of space-time driving trace Forecasting Methodology of blacklist vehicle comprises that concrete steps are as follows to the prediction of blacklist vehicle running path:

Step 1: obtain original traffic data by existing Intellective traffic information system, and be stored in the database;

Step 2: the number-plate number of the vehicle that catching in real time passes through respectively monitors bayonet socket, and compare with the number-plate number on the blacklist, if exist with blacklist on the identical number-plate number, confirm that then this vehicle is the blacklist vehicle, and the current data of the history of in database, searching this blacklist vehicle, execution in step 3; Otherwise abandon the number-plate number of this vehicle, finish;

Step 3: the monitoring bayonet socket that occurs take this blacklist vehicle is as starting point bayonet socket G ₀, cross data mining technology according to the historical current data communication device of this blacklist vehicle of searching in the step 2 and predict that this blacklist vehicle is at the 1st section prediction stroke L ₁In will travel to terminal point bayonet socket G ₁

Step 4: with the prediction this blacklist vehicle at i(i=1,2 ..., m) section prediction stroke L _iIn terminal point bayonet socket G _i(i=1,2 ..., be that this blacklist vehicle is at i+1 section prediction stroke L m) _I+1In the starting point bayonet socket, predict that according to method described in the step 3 this blacklist vehicle is at i+1 section prediction stroke L _I+1In terminal point bayonet socket G _I+1Wherein, each section of this blacklist vehicle prediction stroke L _iSum is the prediction driving path L of this blacklist vehicle, and m is that prediction driving path L comprises the number of predicting stroke;

Step 5: the prediction driving path L that judges this blacklist vehicle comprises the number m that predicts stroke and whether reaches predefined threshold value, if do not reach threshold value, then returns step 4; If reach threshold value, then show the prediction of the prediction driving path L of this blacklist vehicle is finished, finish.

The invention has the beneficial effects as follows: the prediction algorithm that the present invention adopts is based on path space track and the time locus that the forecast model predict future of historical data similarity travels, and current traffic congestion situation regular with reference to the historical track of human pilot improved the accuracy of road prediction probability and the precision of time prediction simultaneously; According to the space-time characteristic of the historical driving trace of blacklist vehicle is predicted this blacklist vehicle driving path in future, and the real-time traffic signatures to predict running time of vehicle pass-through information analysis by road bayonet socket real-time grasp shoot and identification, greatly improved the precision of prediction of vehicle driving trace, this method has good real-time, adaptivity, extensibility.

On the basis of technique scheme, the present invention can also do following improvement.

Further, this blacklist vehicle of prediction is predicted stroke L at the 1st section in the described step 3 ₁In will travel to terminal point bayonet socket G ₁, and predict this blacklist vehicle in the step 4 at i+1(i=1,2 ..., m) section prediction stroke L _I+1(i=1,2 ..., will travel in m) to terminal point bayonet socket G _I+1(i=1,2 ..., concrete steps m) are as follows:

Step 34.1: the historical running data recording of this blacklist vehicle of finding in the step 2 is sorted according to time order and function, to go up two adjacent bayonet sockets and described each self-corresponding elapsed time of two bayonet sockets the time as a stroke recording, this blacklist vehicle elapsed time the preceding bayonet socket as starting point, this blacklist vehicle elapsed time after bayonet socket as terminal point, generate the stroke recording S set of this blacklist vehicle;

Step 34.2: the stroke of this blacklist vehicle in the stroke recording S set is classified according to starting point, and the stroke that will have identical starting point is placed on same candidate's stroke subset S _K(K=A, B, C ...) in;

Step 34.3: take out i+1 with this blacklist vehicle (i=0,1,2 ..., m) section prediction stroke L _I+1(i=0,1,2 ..., starting point bayonet socket G m) _iCandidate's stroke subset S for starting point _K(K=A, B, C ...);

Step 34.4: the date of note current time is W, and the moment of current time is T, and the date of judging the current time is whether W is holiday, if it is from the stroke of candidate described in the step 34.3 subset S _KIn to pick out the date be holiday, this blacklist vehicle through moment of starting point bayonet socket and terminal point bayonet socket respectively with the poor stroke recording less than schedule time Ts of the hours of current time T, put into similar stroke S set _SIn; Otherwise from candidate's stroke subset S _KIn to pick out the date be working day, this blacklist vehicle through moment of starting point bayonet socket and terminal point bayonet socket respectively with the poor stroke recording less than schedule time Ts of the hours of current time T, and put into similar stroke S set _SIn;

Step 34.5: with similar stroke S set _SIn all terminal point bayonet socket and corresponding transit time thereof put into next stop candidate's bayonet socket S set as a record _NIn;

Step 34.6: statistics next stop candidate's bayonet socket S set _NIn the frequency that occurs of each terminal point bayonet socket, select the highest bayonet socket of frequency as this blacklist vehicle of prediction at i+1 section prediction stroke L _I+1In will travel to the terminal point bayonet socket, note is G _I+1

Adopt the beneficial effect of above-mentioned further scheme to be: to utilize the forecast model of historical data similarity, and by the capable form of set related data is processed, greatly improved logic and the arithmetic speed processed.

Further, technique scheme also comprises passes through i(i=1 in the step 4 to this blacklist Vehicle Driving Cycle, and 2 ..., m) section prediction stroke L _iRequired time t _iPrediction, concrete steps are as follows:

Step 4.1: from similar stroke S set _SIn pick out stroke and be prediction stroke L _iAll records, calculate this blacklist vehicle and travel through prediction stroke L at every turn _iRequired time t (i) ';

Step 4.2: calculate this blacklist Vehicle Driving Cycle through i section prediction stroke L _iRequired averaging time

Step 4.3: in database, search apart from the current time in schedule time T2, successively appear at i section prediction stroke L _iStarting point bayonet socket G ₁With terminal point bayonet socket G _I+1All strokes of all vehicles, calculate each vehicle and travel through prediction stroke L at every turn _iTime t (i) ' ';

Step 4.4: all Vehicle Driving Cycles are through predicted path L described in the calculation procedure 4.3 _iAveraging time

Step 4.5: this blacklist Vehicle Driving Cycle in the calculation procedure 4.2 is through i section prediction stroke L _iRequired averaging time

Predict stroke L with described all Vehicle Driving Cycles in the step 4.4 through the i section _iRequired averaging time Weighted mean value, will

With

Weighted mean value as this blacklist Vehicle Driving Cycle through i section prediction stroke L _iPredicted time t _i

Adopt the beneficial effect of above-mentioned further scheme to be: to calculate this blacklist vehicle required averaging time on this section stroke by the historical running data of blacklist vehicle, calculate the averaging time of other vehicles on this section stroke by in database, searching running data apart from current time other vehicles in schedule time T2, this can provide reference frame for the traffic of current road, and by the blacklist vehicle capable in the past averaging time required on this section stroke and in the recent period other vehicles on this section stroke averaging time Accurate Prediction should and the predicted time of list vehicle on this section stroke, greatly improved precision of prediction.

Further, technique scheme also comprises the prediction of this blacklist Vehicle Driving Cycle through the required time t of the driving path of prediction described in the step 5 L, calculates according to following formula:

t=t ₁+t ₂+…t _i+…t _m(i=1,2,…,m)

Wherein, t _iFor this blacklist Vehicle Driving Cycle is predicted stroke L through the i section _iPredicted time.

Adopt the beneficial effect of technique scheme to be: can predict the whole predicted path required time of blacklist vehicle.

Description of drawings

Fig. 1 is the process flow diagram of prediction blacklist vehicle running path of the present invention;

Fig. 2 is step 3 of the present invention, 4 process flow diagram;

Fig. 3 is the process flow diagram of prediction blacklist vehicle of the present invention running time on the driving path of prediction.

Embodiment

Below in conjunction with accompanying drawing principle of the present invention and feature are described, institute gives an actual example and only is used for explaining the present invention, is not be used to limiting scope of the present invention.

As shown in Figure 1, a kind of space-time driving trace Forecasting Methodology of blacklist vehicle comprises that concrete steps are as follows to the prediction of blacklist vehicle running path:

Step 1: obtain original traffic data by existing Intellective traffic information system, and be stored in the database;

Step 2: the number-plate number of the vehicle that catching in real time passes through respectively monitors bayonet socket, and compare with the number-plate number on the blacklist, if exist with blacklist on the identical number-plate number, confirm that then this vehicle is the blacklist vehicle, and the current data of the history of in database, searching this blacklist vehicle, execution in step 3; Otherwise abandon the number-plate number of this vehicle, finish;

Step 3: the monitoring bayonet socket that occurs take this blacklist vehicle is as starting point bayonet socket G ₀, cross data mining technology according to the historical current data communication device of this blacklist vehicle of searching in the step 2 and predict that this blacklist vehicle is at the 1st section prediction stroke L ₁In will travel to terminal point bayonet socket G ₁

Step 4: with the prediction this blacklist vehicle at i(i=1,2 ..., m) section prediction stroke L _iIn terminal point bayonet socket G _i(i=1,2 ..., be that this blacklist vehicle is at i+1 section prediction stroke L m) _I+1In the starting point bayonet socket, predict that according to method described in the step 3 this blacklist vehicle is at i+1 section prediction stroke L _I+1In terminal point bayonet socket G _I+1Wherein, each section of this blacklist vehicle prediction stroke L _iSum is the prediction driving path L of this blacklist vehicle, and m is that prediction driving path L comprises the number of predicting stroke;

Step 5: the prediction driving path L that judges this blacklist vehicle comprises the number m that predicts stroke and whether reaches predefined threshold value, if do not reach threshold value, then returns step 4; If reach threshold value, then show the prediction of the prediction driving path L of this blacklist vehicle is finished, finish.

As shown in Figure 2, this blacklist vehicle of prediction is predicted stroke L at the 1st section in the described step 3 ₁In will travel to terminal point bayonet socket G ₁, and predict this blacklist vehicle in the step 4 at i+1(i=1,2 ..., m) section prediction stroke L _I+1(i=1,2 ..., will travel in m) to terminal point bayonet socket G _I+1(i=1,2 ..., concrete steps m) are as follows:

Step 34.1: the historical running data recording of this blacklist vehicle of finding in the step 2 is sorted according to time order and function, to go up two adjacent bayonet sockets and described each self-corresponding elapsed time of two bayonet sockets the time as a stroke recording, this blacklist vehicle elapsed time the preceding bayonet socket as starting point, this blacklist vehicle elapsed time after bayonet socket as terminal point, generate the stroke recording S set of this blacklist vehicle;

Step 34.2: the stroke of this blacklist vehicle in the stroke recording S set is classified according to starting point, and the stroke that will have identical starting point is placed on same candidate's stroke subset S _K(K=A, B, C ...) in;

Step 34.3: take out i+1 with this blacklist vehicle (i=0,1,2 ..., m) section prediction stroke L _I+1(i=0,1,2 ..., starting point bayonet socket G m) _iCandidate's stroke subset S for starting point _K(K=A, B, C ...);

Step 34.4: the date of note current time is W, and the moment of current time is T, and the date of judging the current time is whether W is holiday, if it is from the stroke of candidate described in the step 34.3 subset S _KIn to pick out the date be holiday, this blacklist vehicle through moment of starting point bayonet socket and terminal point bayonet socket respectively with the poor stroke recording less than schedule time Ts of the hours of current time T, put into similar stroke S set _SIn; Otherwise from candidate's stroke subset S _KIn to pick out the date be working day, this blacklist vehicle through moment of starting point bayonet socket and terminal point bayonet socket respectively with the poor stroke recording less than schedule time Ts of the hours of current time T, and put into similar stroke S set _SIn;

Step 34.5: with similar stroke S set _SIn all terminal point bayonet socket and corresponding transit time thereof put into next stop candidate's bayonet socket S set as a record _NIn;

Step 34.6: statistics next stop candidate's bayonet socket S set _NIn the frequency that occurs of each terminal point bayonet socket, select the highest bayonet socket of frequency as this blacklist vehicle of prediction at i+1 section prediction stroke L _I+1In will travel to the terminal point bayonet socket, note is G _I+1

As shown in Figure 3, to i(i=1 in this blacklist Vehicle Driving Cycle process step 4,2 ..., m) section prediction stroke L _iRequired time t _iPrediction, concrete steps are as follows:

Step 4.1: from similar stroke S set _SIn pick out stroke and be prediction stroke L _iAll records, calculate this blacklist vehicle and travel through prediction stroke L at every turn _iRequired time t (i) ';

Step 4.2: calculate this blacklist Vehicle Driving Cycle through i section prediction stroke L _iRequired averaging time

Step 4.3: in database, search apart from the current time in schedule time T2, successively appear at i section prediction stroke L _iStarting point bayonet socket G ₁With terminal point bayonet socket G _I+1All strokes of all vehicles, calculate each vehicle and travel through prediction stroke L at every turn _iTime t (i) ' ';

Step 4.4: all Vehicle Driving Cycles are through predicted path L described in the calculation procedure 4.3 _iAveraging time

Step 4.5: this blacklist Vehicle Driving Cycle in the calculation procedure 4.2 is through i section prediction stroke L _iRequired averaging time

Predict stroke L with described all Vehicle Driving Cycles in the step 4.4 through the i section _iRequired averaging time

Weighted mean value, will

With Weighted mean value as this blacklist Vehicle Driving Cycle through i section prediction stroke L _iPredicted time t _i

Further, technique scheme also comprises the prediction of this blacklist Vehicle Driving Cycle through the required time t of the driving path of prediction described in the step 5 L, calculates according to following formula:

t=t ₁+t ₂+…t _i+…t _m(i=1,2,…,m)

Wherein, t _iFor this blacklist Vehicle Driving Cycle is predicted stroke L through the i section _iPredicted time.

Embodiment:

Blacklist vehicle driving trace Forecasting Methodology based on the data mining of road bayonet socket information of vehicles comprises the prediction of blacklist vehicle running path and running time prediction.

One, blacklist vehicle running path prediction

1. obtain original traffic data by existing Intellective traffic information system, and be stored in the database;

When the blacklist vehicle for example the 5E097 of A-grade in the first class appear at certain monitoring during bayonet socket, remember that this bayonet socket is bayonet socket G ₀, the current data of history of in database, searching this blacklist vehicle;

3. according to the time order and function order this blacklist vehicle is passed through the current data sorting of history of road bayonet socket, the current data record format of the history after the ordering such as table 1:

Table 1

Record number	Number plate of vehicle	The bayonet socket title	Traveling lane	Travel direction	Elapsed time
						7716873	The 5E097 of A-grade in the first class	Bayonet socket A	1	From west to east	t1,2012-12-0513:00:00
7717228	The 5E097 of A-grade in the first class	Bayonet socket B	1	From west to east	t2,2012-12-0513:03:00
						7716875	The 5E097 of A-grade in the first class	Bayonet socket C	1	The direction of going out of the city	t3,2012-12-0513:06:00
7716869	The 5E097 of A-grade in the first class	Bayonet socket A	1	From west to east	t4,2012-12-0714:08:00
						7717220	The 5E097 of A-grade in the first class	Bayonet socket D	1	By north orientation south	t5,2012-12-0714:14:00
7717311	The 5E097 of A-grade in the first class	Bayonet socket E	1	From south to north	t6,2012-12-0714:18:01
						7717309	The 5E097 of A-grade in the first class	Bayonet socket B	1	From west to east	t7,2012-12-0813:24:01
7717313	The 5E097 of A-grade in the first class	Bayonet socket C	1	The direction of going out of the city	t8,2012-12-0813:29:02
						7716880	The 5E097 of A-grade in the first class	Bayonet socket A	1	By north orientation south	t9,2012-12-0915:32:02
7716889	The 5E097 of A-grade in the first class	Bayonet socket B	1	From west to east	t10,2012-12-0915:34:02
						7716884	The 5E097 of A-grade in the first class	Bayonet socket C	1	The direction of going out of the city	t11,2012-12-0915:40:02
7716943	The 5E097 of A-grade in the first class	Bayonet socket A	1	From west to east	t12,2012-12-1617:22:02
						7716945	The 5E097 of A-grade in the first class	Bayonet socket B	1	From west to east	t13,2012-12-1617:24:43
7716963	The 5E097 of A-grade in the first class	Bayonet socket C	1	The direction of going out of the city	t14,2012-12-1617:30:14
						7716987	The 5E097 of A-grade in the first class	Bayonet socket F	1	The direction of going out of the city	t15,2012-12-1617:37:23

Will the time upper adjacent two bayonet sockets and corresponding elapsed time thereof be as a stroke recording, this blacklist vehicle elapsed time the preceding bayonet socket as start of a run, this blacklist vehicle elapsed time after bayonet socket as the terminal point terminal point.If this blacklist vehicle comprises the n bar by the current data of history of road bayonet socket, n=15 in the present embodiment, then the stroke recording sum is the n-1 bar.Generate the stroke recording S set of this blacklist vehicle:

S={S1=(A,B,t1,t2),S2=（B，C，t2,t3）…S14=(C,F,t14,t15)}；

Be S1=(A, B, t1, t2), S2=(B, C, t2, t3), S3=(C, A, t3, t4), S4=(A, D, t4, t5) S5=(D, E, t5, t6), S6=(E, B, t6, t7), S7=(B, C, t7, t8) and, S8=(C, A, t8, t9), S9=(A, B, t9, t10) and, S10=(B, C, t10, t11), S11=(C, A, t11, t12), S12=(A, B, t12, t13), S13=(B, C, t13, t14), S14=(C, F, t14, t15).

4. the stroke of this blacklist vehicle in the stroke recording S set is classified according to the starting point bayonet socket, the stroke that will have the starting point of identical bayonet socket is placed on a stroke subset S _K(K=A, B, C ...) in, then the subset take bayonet socket K as starting point is calculated as follows,

S _A={S1,S4，S9,S12}；

S _B={S2，S7，S10,S13}；

S _C={S3，S8,S11,S14}；

S _D={S5}；

S _E={S6}。

5. suppose that current blacklist vehicle appears at bayonet socket A, then with bayonet socket A as initial point, namely the starting point bayonet socket of the 1st section prediction stroke L1 is designated as G ₀

6. from the stroke subset described in the step 4, take out with the stroke subset S of bayonet socket A as starting point _A

7. the date of note current time is W, and the moment of current time is T, and the date of judging the current time is whether W is holiday, judges whether the date W of current time is holiday, if it is from the subset of stroke described in the step 6 S _AIn to pick out the date be holiday, poor (Ts is taken as 3 hours less than schedule time Ts with the hours of current time T respectively through moment of starting point bayonet socket and terminal point bayonet socket for this blacklist vehicle, also can adjust as required) stroke recording, put into similar stroke S set _SIn (for example current time T is point 1 day 15 May in 2013, then hours all records between 12 o'clock to 18 o'clock holiday is added similar stroke S set _SOtherwise from candidate's stroke S set _AIn to pick out the date be working day, poor (Ts is taken as 3 hours less than schedule time Ts with the hours of current time T respectively through moment of starting point bayonet socket and terminal point bayonet socket for this blacklist vehicle, also can adjust as required) stroke recording, put into similar stroke S set _SIn;

Suppose that current date is holiday, the data 2012-12-07 in the tables of data 1,2012-12-08, this three dates of 2012-12-09 are the holiday of S then _S={ S4, S9, S12}

8. with similar stroke S set _SIn all terminal point bayonet socket and corresponding transit time put into next stop candidate's bayonet socket S set as a record record _NIn;

S _N={(D,t5),（B，t10）,(B,t13)}

9. add up next stop candidate's bayonet socket S set _NIn the frequency that occurs of each terminal point bayonet socket, select the highest bayonet socket B of frequency as with initial point G in the present embodiment ₀Prediction stroke end bayonet socket G for starting point ₁; In the present embodiment the data bulk of giving limited, data data of giving in the table 1 in the database of reality.

10. the driving path of this blacklist vehicle prediction not only comprises a prediction stroke L ₁, can predict m prediction stroke according to customer requirements, predict that namely the blacklist vehicle will travel through (L future successively ₁, L ₂..., L _m) the section stroke.Terminal point bayonet socket G with prediction in the step 9 ₁(bayonet socket B) is as prediction stroke L ₂Starting point, according to step 6 to the described method of step 9 prediction L ₂Terminal point bayonet socket G ₂, go down successively, until the path L length of prediction meets the requirements of threshold value.If the threshold value of the number m of requirement forecast stroke is 3, then the bayonet socket sequence of predicted path is (G ₀, G ₁, G ₂, G ₃), the driving path L of final prediction is (L ₁, L ₂, L ₃), L wherein ₁Bayonet socket G ₀To bayonet socket G ₁The prediction stroke, L ₂Bayonet socket G ₁To bayonet socket G ₂The prediction stroke, L ₃Bayonet socket G ₂To bayonet socket G ₃The highway section.

Two, the prediction of blacklist vehicle running time on the driving path of prediction

1. predict that this blacklist Vehicle Driving Cycle is through i section required time t _i

With prediction first paragraph stroke L ₁Used time t ₁For example describes, with similar stroke S set _SIn with predicted path in the 1st section prediction stroke L ₁(bayonet socket A-〉bayonet socket B) identical stroke is picked out, namely from bayonet socket G ₀(A) to bayonet socket G ₁(B) stroke comprises S1, and S9 and S12 calculate stroke S1, the used time of S9 and S12, and calculate averaging time of above-mentioned time, be designated as

2. in real-time database, retrieval is generally 1 hour apart from the current time at schedule time T2(T2, for example current time T is 15: 20 on the 1st May in 2013, record before then searching in 1 hour, on May 1st, 1 14: 20 records of assigning between the current time T) in, all Vehicle Driving Cycles are through prediction stroke highway section L ₁Stroke recording, calculate each vehicle through L ₁Required time calculates each car again through L ₁Averaging time, calculate again all vehicle pass-throughs through prediction stroke highway section L ₁Needed averaging time

L wherein ₁The starting point bayonet socket be G ₀(being bayonet socket A in the present embodiment), the terminal point bayonet socket is G ₁(being bayonet socket B in the present embodiment).

3: this blacklist Vehicle Driving Cycle in the calculation procedure 1 is through the 1st section prediction stroke L ₁Required averaging time

In current real-time road section traffic volume situation, travel through i section prediction stroke L with described all vehicles in the step 2 ₁Required averaging time

Weighted mean value, with this weighted mean time as this blacklist Vehicle Driving Cycle through the 1st section prediction stroke L ₁Predicted time t ₁,

$t_1=a\·\stackrel{\‾}{t{\left(1\right)}^{\′}}+(1-a)\·\stackrel{\‾}{t{\left(1\right)}^{\′\′}},0\≤a\≤1$

This blacklist Vehicle Driving Cycle is through other prediction strokes L _iRequired time t _iMethod the same.

Wherein, to the prediction of this blacklist Vehicle Driving Cycle through the required time t of prediction driving path L, calculate according to following formula:

t=t ₁+t ₂+…t _i+…t _m(i=1,2,…,m)

If the threshold value of the number m of the prediction stroke that the path L of prediction comprises is 3, then the driving path L of prediction is (G ₀, G ₁, G ₂, G ₃), G then ₀To G ₁The predicted time that travels is t ₁, G ₁To G ₂The predicted time that travels is t ₂, G ₂To G ₃The predicted time that travels is t ₃Suppose that this blacklist vehicle is at bayonet socket G ₀The time that occurs is T, and then final prediction arrives respectively bayonet socket G ₁, G ₂, G ₃Time be respectively T+t ₁, T+t ₁+ t ₂, T+t ₁+ t ₂+ t ₃

Described road bayonet socket information of vehicles is by the vehicle snapshot camera device collection that is deployed in the urban road bayonet socket, and information of vehicles comprises all transit times by vehicle, captures the number-plate number of photo, Intelligent Recognition etc.Described blacklist vehicle driving trace forecasting techniques comprises based on the driving path of historical track similarity selects Forecasting Methodology, and based on the running time Forecasting Methodology of road traffic condition assessment.Described driving trace prediction algorithm efficient is high, can the rear in real time track of anticipation Vehicle Driving Cycle appear at the blacklist vehicle, determine bayonet socket that this vehicle next one may occur and the time of appearance, in advance these bayonet sockets are carried out video monitoring and deploy to ensure effective monitoring and control of illegal activities, arresting for the blacklist vehicle providing decision-making auxiliary.

The above only is preferred embodiment of the present invention, and is in order to limit the present invention, within the spirit and principles in the present invention not all, any modification of doing, is equal to replacement, improvement etc., all should be included within protection scope of the present invention.

Claims (4)

1. the space-time driving trace Forecasting Methodology of a blacklist vehicle is characterized in that, comprises that concrete steps are as follows to the prediction of blacklist vehicle running path:

Step 1: obtain original traffic data by existing Intellective traffic information system, and be stored in the database;

Step 2: the number-plate number of the vehicle that catching in real time passes through respectively monitors bayonet socket, and compare with the number-plate number on the blacklist, if exist with blacklist on the identical number-plate number, confirm that then this vehicle is the blacklist vehicle, and the current data of the history of in database, searching this blacklist vehicle, execution in step 3; Otherwise abandon the number-plate number of this vehicle, finish;

Step 3: the monitoring bayonet socket that occurs take this blacklist vehicle is as starting point bayonet socket G ₀, cross data mining technology according to the historical current data communication device of this blacklist vehicle of searching in the step 2 and predict that this blacklist vehicle is at the 1st section prediction stroke L ₁In will travel to terminal point bayonet socket G ₁

Step 4: with the prediction this blacklist vehicle at i(i=1,2 ..., m) section prediction stroke L _iIn terminal point bayonet socket G _i(i=1,2 ..., be that this blacklist vehicle is at i+1 section prediction stroke L m) _I+1In the starting point bayonet socket, predict that according to method described in the step 3 this blacklist vehicle is at i+1 section prediction stroke L _I+1In terminal point bayonet socket G _I+1Wherein, each section of this blacklist vehicle prediction stroke L _iSum is the prediction driving path L of this blacklist vehicle, and m is that prediction driving path L comprises the number of predicting stroke;

Step 5: the prediction driving path L that judges this blacklist vehicle comprises the number m that predicts stroke and whether reaches predefined threshold value, if do not reach threshold value, then returns step 4; If reach threshold value, then show the prediction of the prediction driving path L of this blacklist vehicle is finished, finish.

2. the space-time driving trace Forecasting Methodology of described a kind of blacklist vehicle according to claim 1 is characterized in that, this blacklist vehicle of prediction is at the 1st section prediction stroke L in the step 3 ₁In will travel to terminal point bayonet socket G ₁, and predict this blacklist vehicle in the step 4 at i+1(i=1,2 ..., m) section prediction stroke L _I+1(i=1,2 ..., will travel in m) to terminal point bayonet socket G _I+1(i=1,2 ..., concrete steps m) are as follows:

Step 34.1: the historical running data recording of this blacklist vehicle of finding in the step 2 is sorted according to time order and function, to go up two adjacent bayonet sockets and described each self-corresponding elapsed time of two bayonet sockets the time as a stroke recording, this blacklist vehicle elapsed time the preceding bayonet socket as starting point, this blacklist vehicle elapsed time after bayonet socket as terminal point, generate the stroke recording S set of this blacklist vehicle;

Step 34.2: the stroke of this blacklist vehicle in the stroke recording S set is classified according to starting point, and the stroke that will have identical starting point is placed on same candidate's stroke subset S _K(K=A, B, C ...) in;

Step 34.3: take out i+1 with this blacklist vehicle (i=0,1,2 ..., m) section prediction stroke L _I+1(i=0,1,2 ..., starting point bayonet socket G m) _iCandidate's stroke subset S for starting point _K(K=A, B, C ...);

Step 34.4: the date of note current time is W, and the moment of current time is T, and the date of judging the current time is whether W is holiday, if it is from the stroke of candidate described in the step 34.3 subset S _KIn to pick out the date be holiday, this blacklist vehicle through moment of starting point bayonet socket and terminal point bayonet socket respectively with the poor stroke recording less than schedule time Ts of the hours of current time T, put into similar stroke S set _SIn; Otherwise from candidate's stroke subset S _KIn to pick out the date be working day, this blacklist vehicle through moment of starting point bayonet socket and terminal point bayonet socket respectively with the poor stroke recording less than schedule time Ts of the hours of current time T, and put into similar stroke S set _SIn;

Step 34.5: with similar stroke S set _SIn all terminal point bayonet socket and corresponding transit time thereof put into next stop candidate's bayonet socket S set as a record _NIn;

Step 34.6: statistics next stop candidate's bayonet socket S set _NIn the frequency that occurs of each terminal point bayonet socket, select the highest bayonet socket of frequency as this blacklist vehicle of prediction at i+1 section prediction stroke L _I+1In will travel to the terminal point bayonet socket, note is G _I+1

3. the space-time driving trace Forecasting Methodology of described a kind of blacklist vehicle according to claim 1 is characterized in that, also comprise this blacklist Vehicle Driving Cycle through i(i=1 in the step 4, and 2 ..., m) section prediction stroke L _iRequired time t _iPrediction, concrete steps are as follows:

Step 4.1: from similar stroke S set _SIn pick out stroke and be prediction stroke L _iAll records, calculate this blacklist vehicle and travel through prediction stroke L at every turn _iRequired time t (i) ';

Step 4.2: calculate this blacklist Vehicle Driving Cycle through i section prediction stroke L _iRequired averaging time

Step 4.3: in database, search apart from the current time in schedule time T2, successively appear at i section prediction stroke L _iStarting point bayonet socket G ₁With terminal point bayonet socket G _I+1All strokes of all vehicles, calculate each vehicle and travel through prediction stroke L at every turn _iTime t (i) ' ';

Step 4.4: all Vehicle Driving Cycles are through predicted path L described in the calculation procedure 4.3 _iAveraging time

Step 4.5: this blacklist Vehicle Driving Cycle in the calculation procedure 4.2 is through i section prediction stroke L _iRequired averaging time

Predict stroke L with described all Vehicle Driving Cycles in the step 4.4 through the i section _iRequired averaging time Weighted mean value, will

With

Weighted mean value as this blacklist Vehicle Driving Cycle through i section prediction stroke L _iPredicted time t _i

4. the space-time driving trace Forecasting Methodology of described a kind of blacklist vehicle according to claim 1 is characterized in that, also comprises the prediction of this blacklist Vehicle Driving Cycle through the required time t of the driving path of prediction described in the step 5 L, calculates according to following formula:

t=t ₁+t ₂+…t _i+…t _m(i=1,2,…,m)

Wherein, t _iFor this blacklist Vehicle Driving Cycle is predicted stroke L through the i section _iPredicted time.

Images