CN102622880A - Traffic information data recovery method and device - Google Patents

Abstract

The invention discloses a traffic information data recovery method and a traffic information data recovery device, and relates to a traffic information technology. In the embodiment of the invention, firstly, a related road with large relevancy with a target road is determined and then a regression coefficient is determined according to historical traffic information data of the target road and the related road, so that when real-time data of the target road is missed or is inaccurate, traffic information data of the target road in a current time period can be determined by the traffic information data of the related road with the largest relevancy and the regression coefficient of the target road and the related roads; and the traffic information data of the target road is determined by the mode, so that the filling of real-time traffic flow missing data and the correction of the real-time data can be implemented and the integrity and the accuracy of traffic information publishing are also improved.

Description

A kind of traffic information data restorative procedure and device

Technical field

The present invention relates to the transport information technology, relate in particular to a kind of traffic information data restorative procedure and device.

Background technology

Follow quickening of urbanization process, the raising of income level of resident, motor vehicle quantity significantly increases, and causes the urban traffic blocking situation increasingly serious, becomes the difficult problem of puzzlement urban development.

Traffic-information service is as the part in advanced person's the traffic information system (ATIS), becomes present transport solution one of the main method of problem of blocking up.Transport information ageingly is divided into two types according to content distributed, and one type is static traffic information, and another kind of is dynamic information.Because static traffic information can't provide service accurately and fast for traveler, is difficult to satisfy user's real-time demand for services.The dynamic real-time traffic information service is for ageing the having relatively high expectations of information issue; The information in the moment on the same day can be provided for the driver; Realized the real-time release of transport information, made the driver can understand the road conditions in the place ahead more directly, more quickly, so that in time adjust the walking along the street line.

Real-time traffic stream information acquisition mode has a variety of; Mainly contain mobile phone, Floating Car, microwave detector, video detector, earth coil etc. at present; Can obtain the traffic parameter such as average velocity, hourage, queue length of road traffic flow through checkout equipment, through further calculating can obtain the transport information that is used to issue to traffic parameter.

But no matter which kind of acquisition mode in gatherer process owing to reasons such as buildings blocks, communication short trouble, checkout equipment chance failures; Inevitably can exist the data deficient phenomena to exist; And the integrality that covers that releases news is the important indicator of evaluating traffic information service quality with the issue accuracy, and therefore repairing for information service provider to missing data is a very important job.Simultaneously, when in the real-time image data of certain bar road because sample size very little or the interference that receives other factors when causing collection value accuracy to reduce, also need be carried out the correction of data.

The method that the data that at present appearance in the transport information lacked are carried out compensation data is mainly, and the transport information according to real-time processing generates is optimized road network; Transport information to real-time processing generates is carried out the abnormal data rejecting, obtains meeting the data of current traffic tendency; From historical data base, choose the historical data that meets this traffic tendency variation according to the real-time traffic trend data that obtains, as the ancillary sources of absence information compensation; According to ancillary sources and road network through optimizing; With a vacancy road chain road is the center construction compensatory zone; Being unit with the zone compensates the road chain that does not have driving trace information in the road network and cover; Confirm to fill up pattern according to vacancy road chain quantity in the road chain role of vacancy road chain and the compensatory zone, thereby accomplish compensation.

This method compensates missing data through the variation tendency of traffic parameter; But; The traffic parameter variation tendency has only been represented the time dependent plots changes of traffic parameter; The similar situation of traffic tendency when the value of two road traffic parameter differs greatly, also can occur, use the data of these roads to compensate, the offset that obtains and the gap of true value are bigger.

Summary of the invention

The embodiment of the invention provides a kind of traffic information data restorative procedure and device, to realize the correction of the filling up of real-time traffic stream missing data, real time data, improves the integrality and the accuracy of transport information issue.

A kind of traffic information data restorative procedure comprises:

Confirm traffic information data with the relevant road of target road correlativity maximum;

According to the regression coefficient between said target road and the said relevant road; And the current traffic information data of said relevant road; Confirm the traffic information data that said target road is current, said regression coefficient is confirmed with the historical traffic information data of said relevant road according to said target road.

A kind of traffic information data prosthetic device comprises:

First data are confirmed the unit, are used for confirming the traffic information data with the relevant road of target road correlativity maximum;

Second data are confirmed the unit; Be used for according to the regression coefficient between said target road and the said relevant road; And the current traffic information data of said relevant road; Confirm the traffic information data that said target road is current, said regression coefficient is confirmed with the historical traffic information data of said relevant road according to said target road.

The embodiment of the invention provides a kind of traffic information data restorative procedure and device; Confirm the relevant road bigger with the target road correlativity; Confirm regression coefficient according to target road with the historical traffic information data of relevant road; When target road real time data disappearance or real time data are inaccurate; Traffic information data that can be through the maximum relevant road of correlativity and target road and the regression coefficient of this relevant road are confirmed the traffic information data of target road; Confirm the traffic information data of current period of target road through this mode, can realize the correction of the filling up of real-time traffic stream missing data, real time data, improve the integrality and the accuracy of transport information issue simultaneously.

Description of drawings

The traffic information data restorative procedure process flow diagram 1 that Fig. 1 provides for the embodiment of the invention;

Fig. 2 confirms method flow diagram for the regression coefficient that the embodiment of the invention provides;

Fig. 3 sets up process flow diagram for the database that the embodiment of the invention provides;

The data screening process flow diagram that Fig. 4 provides for the embodiment of the invention;

Fig. 5 confirms method flow diagram for the time correlation coefficient that the embodiment of the invention provides;

The cross roads structural representation that Fig. 6 provides for the embodiment of the invention;

Fig. 7 for the embodiment of the invention provide based on temporal correlation tabulation method flow diagram with the definite relevant road of spatial coherence tabulation;

The comparison synoptic diagram of the traffic information data of relevant Traffic Information data of the morning peak period that Fig. 8 provides for the embodiment of the invention and target road;

The morning peak period that Fig. 9 provides for the embodiment of the invention is confirmed the relative error value of target road traffic information data through relevant road;

The comparison synoptic diagram of the traffic information data of relevant Traffic Information data of the whole day period that Figure 10 provides for the embodiment of the invention and target road;

The whole day period that Figure 11 provides for the embodiment of the invention is confirmed the relative error value of target road traffic information data through relevant road;

The relevant road 1 that Figure 12 provides for the embodiment of the invention and the relative error histogram of target road;

The relevant road 2 that Figure 13 provides for the embodiment of the invention and the relative error histogram of target road;

The traffic information data prosthetic device structural representation that Figure 14 provides for the embodiment of the invention.

Embodiment

The embodiment of the invention provides a kind of traffic information data restorative procedure and device; Confirm the relevant road bigger with the target road correlativity; Confirm regression coefficient according to target road with the historical traffic information data of relevant road; When target road real time data disappearance or real time data are inaccurate; Traffic information data that can be through the maximum relevant road of correlativity and target road and the regression coefficient of this relevant road are confirmed the traffic information data of current period of target road; Confirm the traffic information data of target road through this mode, can realize the correction of the filling up of real-time traffic stream missing data, real time data, improve the integrality and the accuracy of transport information issue simultaneously.

As shown in Figure 1, the traffic information data restorative procedure that the embodiment of the invention provides comprises:

Step S101, confirm the traffic information data with the relevant road of target road correlativity maximum;

Step S102, according to the regression coefficient between target road and the relevant road; And the current traffic information data of relevant road; Confirm the traffic information data that target road is current, regression coefficient is confirmed with the historical traffic information data of relevant road according to target road.

The traffic information data restorative procedure that uses the embodiment of the invention to provide needs at first to confirm and the maximum relevant road of target road correlativity.Usually, can pre-determine several relevant roads bigger with the target road correlation, again when needs obtain the target road traffic information data, therefrom select traffic information data accurately and with the maximum relevant road of target road correlation.

When the relevant road of confirming with target road correlation maximum; Can at first confirm road with the maximum setting quantity of the temporal correlation of target road; The tabulation of formation time correlation, wherein, temporal correlation is confirmed based on the speed time series of corresponding period of two road; The road of the setting quantity that spatial coherence definite again and target road is maximum forms the spatial coherence tabulation, and wherein, spatial coherence is confirmed based on the space length and the connectedness in two road; At last, based on temporal correlation tabulation with spatial coherence tabulation confirm and the relevant road of target road correlation maximum.

Concrete, below in conjunction with Fig. 2 the traffic information data restorative procedure that the embodiment of the invention provides is carried out

Specify:

Because the traffic flow time series is tieed up in time dimension and space and existed stronger regularity aspect two, the present invention proposes the real time traffic data restorative procedure.Substitute the transport information that lacks road through seeking than the transport information on rope-race road, realize filling up of missing data with known disappearance road time, spatial coherence; When in the real time data because when gathering sample size deficiency or other reasons and causing the real time data distortion, the transport information of distortion road can utilize this method to find correlativity to come distortion road real time data is proofreaied and correct than the rope-race road.In embodiments of the present invention, road is meant the least unit of transport information issue, certainly, when practical application, also can confirm the unit of road according to actual conditions.Disappearance road and distortion road are referred to as the problem road here, just need to confirm according to relevant road the target road of traffic information data.

As shown in Figure 2, the preferable traffic information data restorative procedure that the embodiment of the invention provides comprises:

Step S201, road in the road network is classified according to category of roads (being expressway, through street, major trunk roads, subsidiary road and branch road) and design speed per hour, make space road network table;

Step S202, set up the traffic flow pattern database through the traffic flow data of a plurality of cycles, period;

Step S203, through to the rejecting of abnormal data in the traffic flow pattern database, be that the temporal correlation analysis is accomplished data and prepared;

Step S204, obtain the time correlation coefficient between the different roads in the road network through the temporal correlation algorithm; At first confirm the included angle cosine similarity coefficient; Be used for estimating the traffic flow time trend in same in cycle in the week same period of the different roads of space road network; Adopt standardized manhatton distance to calculate the distance of each road speeds in the road network of the interior space of each release cycle of day part then, last comprehensive included angle cosine similarity coefficient and two parameters of standardization manhatton distance obtain the time correlation coefficient;

Step S205, according to time correlation coefficient rank order from big to small, the time correlation road is write in the road traffic flow temporal correlation tabulation;

Step S206, according to the spatial coherence between the distance calculation road between the road in the road network, when calculating the relative distance of different roads in target road and the road network, the relative distance that is not communicated with road is set at infinity.Because traffic congestion and traffic events exist contingency and randomness; The accuracy of the transport information that therefore obtains through temporal correlation also can be affected to a certain extent; Need to consider the space correlation relation between the road, the steric interaction of geographical space road network exists with the distance extension the constantly characteristic of decay;

Step S207, to space relative distance according to from small to large rank order, get n before the rank _SpitialThe road of name forms the spatial coherence tabulation of each bar road in the road network as the space correlation road;

Step S208 according to temporal correlation tabulation and spatial coherence tabulation, confirms to set the relevant road of quantity;

Step S209, target road is carried out regretional analysis with relevant road traffic flow time series, set up regression equation, obtain the regression coefficient of target road and relevant road;

Step S210, target road in the road network is write in the traffic flow pattern database with the regression coefficient of relevant road.

In case when disappearance or distortion appear in target road Real-time Traffic Information data, the traffic information data and the regression coefficient in relevant this moment of road are with it multiplied each other, just can obtain the traffic information data in this moment of target road.

The transport information restorative procedure that the embodiment of the invention provides can improve the issue density and the information accuracy of road network information.This method is applicable to different traffic flow data sampling modes, as long as the travelling speed of road network table, road network road can be provided, even when a plurality of cycles of this road all do not have real time data, also can adopt this method to carry out the reparation of data.

In step S201, the space road network table of made can generate on the basis of the electronic chart of having encoded.Carry out the division of road network space attribute according to category of roads and highway layout speed per hour, category of roads comprises expressway, through street, major trunk roads, subsidiary road and branch road etc.Because can there be the design speed per hour condition of different of different roads under the same category of roads in the spatial form of road, radius of turn, geographic position of living in difference, so when carrying out the spatial character fundamental analysis, have considered this factor.Table 1 is a road network table instance.

Table 1 road network table

As shown in Figure 3, in step S202, set up the traffic flow pattern database through the traffic flow data of a plurality of cycles, period, specifically comprise:

The historical traffic information data of step S301, the nearest certain hour section of selection road network advised that the length of selected nearest certain hour section can be set in advance three months;

Step S302, selected data are carried out mark according to category of roads;

Step S303, selected data are carried out mark according to setting cycle, for example this cycle can be week;

Step S304, selected data are carried out mark according to setting the period; Because the period of right time (peak, Ping Feng, festivals or holidays etc.) is bigger to the influence of traffic characteristics, maybe be different in the different periods with the similar road of problem road traffic state, therefore need carry out mark according to the period to selected data;

Step S305, selected data are carried out mark according to weather (rain, snow, mist etc.);

The data of step S306, selection setting cycle, period and weather, and calculate average velocity; For example, the average velocity of road certain peak period cycle in week through nearest a period of time in the same cycle in week should peak period speed obtain through modes such as direct average or weighted means.

When setting up the traffic flow pattern database, several row below in database, increasing: relevant road codes 1, relevant road codes 2, regression coefficient 1, regression coefficient 2, initial value is empty; Be to be example in the embodiment of the invention with two relevant roads; In practical operation, can select a plurality of relevant roads according to actual conditions; Then should increase and corresponding road codes of relevant road quantity and regression coefficient this moment in database.Wherein, regression coefficient is that the relevant road with corresponding encoded of target road carries out regression coefficient and calculates through the correlation calculations method respectively.

It is a kind of that to set up the traffic flow pattern database as shown in table 2:

Table 2 traffic flow pattern database

As shown in Figure 4, in step S203,, accomplish data for the temporal correlation analysis and prepare through rejecting to abnormal data in the traffic flow pattern database, specifically comprise:

Step S401, the period of confirming to participate in analysis, reject the outer data of period that participation is analyzed, for example; If need carry out global analysis to the data on daytime; Can reject the data of 23:00-06:55 period,, then can reject the data of 22:00-16:00 period if only need to analyze the data of evening peak period; In practical operation, can set the time period of rejecting peak period based on the traffic flow of various places.；

Step S402, the link length minimum value of confirming to participate in analysis are rejected the road of mileages of transport route less than this link length minimum value; For example, be 50 meters if confirm the link length minimum value, then reject link length less than 50 meters road;

Step S403, reject and to participate in the period of analyzing, speed is the road that ratio that 0 record quantity accounts for total quantity reaches preset proportion; For example this ratio can be set at 30%-50%, when this ratio is higher, explains that then this road often can not obtain traffic information data, and the traffic information data that perhaps obtains is inaccurate, so reject corresponding road;

Step S404, reject occurrence of traffic accident, hold large-scale activity, the relevant road segments of traffic control and period.

Through step S401-step S404, can determine the conventional traffic information data that can be used for calculating regression coefficient.Certainly, when practical operation, can also further reject according to factors such as weather.

As shown in Figure 5, in step S204, obtain the time correlation coefficient between the different roads in the road network through the temporal correlation algorithm, specifically comprise:

Different roads are with the traffic flow time trend in one-period in step S501, the utilization included angle cosine similarity coefficient evaluation space road network;

The distance of each road speeds in the road network of space in step S502, standardized each release cycle of manhatton distance computational analysis period of employing;

Step S503, comprehensive included angle cosine similarity coefficient and two parameters of standardization manhatton distance obtain the time correlation coefficient.

Concrete, step S501 specifically comprises:

To the road j identical in target road i and the road network, confirm that i, j two road speed seasonal effect in time series included angle cosine similarity coefficient are with target road i category of roads:

$s_{\mathrm{ij}}=\frac{\underset{t=1}{\overset{p}{\mathrm{\Σ}}}{x}_{\mathrm{it}}{x}_{\mathrm{jt}}}{\sqrt{\underset{t=1}{\overset{p}{\mathrm{\Σ}}}{x_{\mathrm{it}}}^2\·\underset{t=1}{\overset{p}{\mathrm{\Σ}}}{x_{\mathrm{jt}}}^2}}(i,j=1......n)---\left(1\right)$

Step S502 specifically comprises: confirm that i, j two road speed seasonal effect in time series manhatton distance are:

$M_{\mathrm{dij}}=\underset{t=1}{\overset{p}{\mathrm{\Σ}}}|x_{\mathrm{it}}-x_{\mathrm{jt}}|(i,j=1......n)---\left(2\right)$

Settling the standard back i, j two road speed seasonal effect in time series standard manhatton distance are:

$R_{\mathrm{dij}}=\frac{\max \left(M_{\mathrm{dik}}\right)-M_{\mathrm{dij}}}{\max \left(M_{\mathrm{dik}}\right)-\min \left(M_{\mathrm{dik}}\right)}---\left(3\right)$

Step S503 specifically comprises: confirm that i, j two road time correlation coefficient are:

C _ij＝ω·R _dij+θ·S _ij (ω+θ＝1，0≤ω≤1，0≤θ≤1) (4)

In formula (1) (2) (3) (4), x _ItBe the speed time series of i bar road t each release cycle of period, x _JtBe the speed time series of road j to be determined at t each release cycle of period, i, j=1 ... N; N is the quantity of each grade road of system-wide net, and its value can obtain t=1 from the road network table ... P; P is the release cycle number of day part; If with 5 minutes be release cycle, if whole day 24 hours all participates in analyzing, p=12 * 24=288 then; M _DikBe other roads and the i bar road traffic flow speed seasonal effect in time series manhatton distance of the same link grade except j bar road in the road network, k=1 ... N, k ≠ i, k ≠ j is if a bar road and i bar road distance are max (M farthest _Dik)=M _Dia, the nearest of b bar road is min (M in i bar road and the road network _Dik)=M _Dib, a ≠ b, R _Dij∈ [0,1]; When a=b, a road is only arranged, be this road self, when basis of calculation manhatton distance, need not calculate the manhatton distance of road self; ω, θ are predefined weights.

Wherein, the t period is set according to traffic conditions and weather condition, is the period of specifically being analyzed, and usually, t was more than or equal to 1 hour.

Similarity coefficient S _IjCharacterized the similarity of i and j two road speed time series variation tendency.As similarity coefficient S _IjApproach 1 more, two road speed time series is relevant more, works as S _IjApproached 0 o'clock, and do not have correlativity between the time series of two road.

Standard manhatton distance R _DijCharacterized i, each release cycle velocity amplitude of speed time series total difference in j two road changes, and works as R _DijApproach 1 more, i, speed time series each release cycle velocity amplitude gap in j two road is more little, approaches 0 more, and gap is big more.

Wherein, C _IjBe the time correlation coefficient that obtains according to after included angle cosine similarity coefficient and the manhatton distance standardization, C _Ij∈ [0,1].Work as C _IjApproach 1 more, i, j two road correlativity is good more, approaches 0 more, and correlativity is poor more.

Since traffic congestion and traffic events have contingency and randomness; The accuracy of transport information also can be affected to a certain extent; Therefore the road correlativity that only relies on traffic flow speed time series to obtain is unsettled, relative dynamic, can not guarantee fully to use this correlation analysis result to carry out the polishing of missing data and the accuracy of the transport information that obtains.

In geography, propose; The steric interaction of space road network exists with distance and extends and the characteristic of continuous decay; This space mutual relationship is static relatively; Have only in the road network that road is newly repaiied, reorganization and expansion etc. causes the road form Shi Caihui that changes to change the space mutual relationship, therefore in this invention, spatial coherence combined with traffic flow speed seasonal effect in time series temporal correlation, confirms the correlativity between the different roads in the road network.

In step S206 and step S207; Then according to the spatial coherence between the road; Formation spatial coherence tabulation, concrete, can be to the road j identical in target road i and the road network with target road i category of roads; Confirm the shortest operating range of road i and road j, confirm that the road of the setting quantity that the shortest operating range is minimum is recorded in the spatial coherence tabulation.

For example, as shown in Figure 6, in this decussation mouth, a road is described to a directed line segment in map, and the starting point of road, terminal point are described to node respectively, like road R among the figure ₁Start node be N ₁, terminal node is N ₂Bee-line between the calculating two road is as the relative distance in two road.Simultaneously, through the topological relation of road in the road network, the relative distance that is not communicated with between the road is made as infinity, like road R ₅At node N ₆Place no left turn, then R ₈, R ₉With R ₅Distance be infinitely great.Except that special circumstances (no through traffic, single file, be not communicated with), the relative distance of road is according to the length computation of wagon flow direction from the road that starting point is passed through of terminal point to another road of a road, i.e. R ₆With R ₅The relative distance computing method be from R ₆Terminal point N ₇To R ₅Starting point N ₇Distance, actual range is 0; R ₇With R ₁The relative distance computing method be from R ₇Terminal point N ₆To node N ₃To R ₁Starting point N ₂Distance, actual range is R ₈And R ₂The length sum, by that analogy.

According to from small to large rank order, when the road that goes wrong was identical with the distance of many roads in the road network, the ordering sequence number was constant with the relative distance of other roads in problem road and the road network.In order to improve counting yield and accuracy rate, we only choose the preceding n of ordering _SpatialThe road of name gets into the tabulation of road spatial coherence.

In step S208, according to temporal correlation tabulation and spatial coherence tabulation, confirm to set the relevant road of quantity, specifically comprise:

Confirm the maximum a bar road of time correlation coefficient in the temporal correlation tabulation, confirm that the b bar road in the spatial coherence tabulation is relevant road in this a bar road, a is predefined value, b≤a;

When a and b are unequal; In other road in the spatial coherence tabulation except that b bar road; Confirm the minimum road of the shortest operating range of a-b bar and target road; When the road travel permit number of determining based on spatial coherence tabulation during, select in tabulation is determined based on spatial coherence the road a-b bar road that the time correlation coefficient is maximum greater than a-b;

In b bar road and a-b bar road, confirm and the maximum relevant road of target road correlativity.

For example; If need to select two relevant roads; Judge at first then that in the tabulation of the temporal correlation of target road maximally related two road are whether in the spatial coherence tabulation; If, think that then these two road are the rope-race road of temporal-spatial fusion correlativity, write in the traffic flow pattern database; If these two road are not in the spatial coherence tabulation; Then need search road relevant with it in the spatial coherence tabulation; Promptly search the most forward road of spatial coherence ordering, if when having identical many roads of distance in the space correlation road, the ordering in the temporal correlation of each comfortable problem road that then relies on them; Determine the rope-race road of the corresponding correlativity of target road, write in the traffic flow pattern database.

Concrete, as shown in Figure 7, when two relevant roads are confirmed in tabulation with spatial coherence through the temporal correlation tabulation, specifically comprise:

Step S701, confirm that according to temporal correlation tabulation rank is at preceding two road;

Step S702, judge determined two road whether all in the spatial coherence tabulation, if, execution in step S703 then, otherwise execution in step S704;

Step S703, confirm that determined two road are the relevant road of target road;

Step S704, confirm in the determined two road that the road in the spatial coherence tabulation is the relevant road of target road, and confirm the required relevant road quantity of further confirming;

Step S705, in spatial coherence tabulation, search the road of the relevant road quantity that needs further to confirm with the strongest being at least of target road correlativity; When many roads are identical with the target road distance; Then, then all be selected into many identical roads of target road distance if need be selected into wherein at least one road;

Step S706, judge the relevant road quantity whether the road number found is further confirmed greater than needs, if, execution in step S707, otherwise execution in step S708;

Step S707, based on temporal correlation tabulation, the road of the relevant road quantity of further confirming with the strongest needs of target road correlation in the road of confirming to be found is relevant road;

Step S708 confirms that the road that is found is relevant road.

In step S209, target road is carried out regretional analysis with relevant road traffic flow time series, set up regression equation, obtain the regression coefficient of target road and relevant road, specifically comprise:

Because many correlativitys are than having multicollinearity between the rope-race road; Therefore the traffic flow time series of a relevant road and target road is carried out regretional analysis; Calculate the regression coefficient in two road; Target road is carried out regretional analysis respectively through relevant road with many, sets up the regression coefficient sequence.

x _it＝α _ij×x _jt+c _it

Wherein, x _ItBe the traffic flow speed time series of target road, x _JtBe relevant road traffic flow speed time series, i, j=1 ... N, n are the road quantity of each grade of system-wide net, and its value can obtain t=1 from the road network table ... P, p are the release cycle numbers of day part, α _IjBe regression coefficient, c _ItBe the constant term of setting, for ease of calculating, usually with c _ItBe made as 0, regression coefficient can write in the table 2 after confirming.

Through the traffic information data restorative procedure that the embodiment of the invention provides, carry out traffic information data when repairing, can realize the correction of the filling up of real-time traffic stream missing data, real time data, improve the integrality and the accuracy of transport information issue simultaneously.

For example; In the morning peak period target road is carried out correlation analysis; Draw in morning peak during the period the two road of relevant road (being relevant road 1 and relevant road 2); The traffic flow time-serial position of two relevant roads and target road when Fig. 8 is morning peak (supposition problem road), the relative error of two relevant roads and target road when Fig. 9 be morning peak, obviously find out the road 1 of being correlated with than the traffic flow variation tendency of relevant road 2 more near the variation tendency of target road.

For example; If select the whole day period that the problem road is carried out correlation analysis; Figure 10 is the traffic flow time-serial position of relevant road of whole day period and target road (supposition problem road); Figure 11 is the traffic flow speed time series relative error of each relevant road of whole day period and target road; Relative error is in-30%～30% ratio and accounts for more than 80% and (see Figure 12, Figure 13), and the whole day average relative error is less than 10%, and relevant road can substitute the reparation that target road is carried out real time data fully.

The embodiment of the invention is also corresponding to provide a kind of traffic information data prosthetic device, shown in figure 14, comprising:

First data are confirmed unit 1401, are used for confirming the traffic information data with the relevant road of target road correlativity maximum;

Second data are confirmed unit 1402; Be used for according to the regression coefficient between target road and the relevant road; And the current traffic information data of relevant road, confirm the traffic information data that target road is current, regression coefficient is confirmed with the historical traffic information data of relevant road according to target road.

Wherein, first data confirm that unit 1401 also is used for:

Confirm and the maximum relevant road of target road correlativity.

Further, first data confirm that unit 1401 is confirmed and the relevant road of target road correlativity maximum, specifically comprise:

The road of the setting quantity that temporal correlation definite and target road is maximum, the tabulation of formation time correlativity, temporal correlation is confirmed according to the speed time series of corresponding period of two road;

The road of the setting quantity that spatial coherence definite and target road is maximum forms the spatial coherence tabulation, and spatial coherence is confirmed based on the space length and the connectedness in two road;

Based on the relevant road that the temporal correlation tabulation is definite with the spatial coherence tabulation and the target road correlation is maximum.

First data confirm that unit 1401 is confirmed and the road of the setting quantity that the temporal correlation of target road is maximum, and the tabulation of formation time correlativity specifically comprises:

To the road j identical in target road i and the road network, confirm that i, j two road speed seasonal effect in time series included angle cosine similarity coefficient are with target road i category of roads:

$s_{\mathrm{ij}}=\frac{\underset{t=1}{\overset{p}{\mathrm{\Σ}}}{x}_{\mathrm{it}}{x}_{\mathrm{jt}}}{\sqrt{\underset{t=1}{\overset{p}{\mathrm{\Σ}}}{x_{\mathrm{it}}}^2\·\underset{t=1}{\overset{p}{\mathrm{\Σ}}}{x_{\mathrm{jt}}}^2}}(i,j=1......n)$

Confirm that i, j two road speed seasonal effect in time series manhatton distance are:

$M_{\mathrm{dij}}=\underset{t=1}{\overset{p}{\mathrm{\Σ}}}|x_{\mathrm{it}}-x_{\mathrm{jt}}|(i,j=1......n)$

Settling the standard back i, j two road speed seasonal effect in time series standard manhatton distance are:

$R_{\mathrm{dij}}=\frac{\max \left(M_{\mathrm{dik}}\right)-M_{\mathrm{dij}}}{\max \left(M_{\mathrm{dik}}\right)-\min \left(M_{\mathrm{dik}}\right)}$

Confirm that i, j two road time correlation coefficient are:

C _ij＝ω·R _dij+θ·S _ij (ω+θ＝1，0≤ω≤1，0≤θ≤1)

To be recorded in the temporal correlation tabulation with the road of the maximum setting quantity of target road i time correlation coefficient;

Wherein, x _ItBe the speed time series of i bar road t each release cycle of period, x _JtBe the speed time series of road j to be determined at t each release cycle of period, i, j=1 ... N, n are the quantity of each grade road of system-wide net, and its value can obtain t=1 from the road network table ... P, p are the release cycle numbers of day part; M _DikBe other roads and the i bar road traffic flow speed seasonal effect in time series manhatton distance of the same link grade except j bar road in the road network, k=1 ... N, k ≠ i, k ≠ j; ω, θ are predefined weights.

First data confirm that unit 1401 is confirmed and the road of the setting quantity that the spatial coherence of target road is maximum, form the spatial coherence tabulation, specifically comprise:

To the road j identical in target road i and the road network with target road i category of roads, confirm the shortest operating range of road i and road j, confirm that the road of the setting quantity that the shortest operating range is minimum is recorded in the spatial coherence tabulation.

First data confirm that unit 1401 according to the relevant road that the temporal correlation tabulation is definite with the spatial coherence tabulation and the target road correlativity is maximum, specifically comprises:

Confirm the maximum a bar road of time correlation coefficient in the temporal correlation tabulation, confirm that the b bar road in the spatial coherence tabulation is relevant road in this a bar road, a is predefined value, b≤a;

When a and b are unequal; In other road in the spatial coherence tabulation except that b bar road; Confirm the minimum road of the shortest operating range of a-b bar and target road; When the road travel permit number of determining based on spatial coherence tabulation during, select in tabulation is determined based on spatial coherence the road a-b bar road that the time correlation coefficient is maximum greater than a-b;

In b bar road and a-b bar road, confirm and the maximum relevant road of target road correlativity.

The embodiment of the invention provides a kind of traffic information data restorative procedure and device; Confirm the relevant road bigger with the target road correlativity; Confirm regression coefficient according to target road with the historical traffic information data of relevant road; When target road data disappearances or data were inaccurate, traffic information data that can be through the maximum relevant road of correlativity and target road and the regression coefficient of this relevant road were confirmed the traffic information data of target road, confirmed the traffic information data of target road through this mode; The correction of the filling up of real-time traffic stream missing data, real time data be can realize, the integrality and the accuracy of transport information issue improved simultaneously.

Those skilled in the art should understand that embodiments of the invention can be provided as method, system or computer program.Therefore, the present invention can adopt the form of the embodiment of complete hardware embodiment, complete software implementation example or combination software and hardware aspect.And the present invention can be employed in the form that one or more computer-usable storage medium (including but not limited to magnetic disk memory, CD-ROM, optical memory etc.) that wherein include computer usable program code go up the computer program of implementing.

The present invention is that reference is described according to the process flow diagram and/or the block scheme of method, equipment (system) and the computer program of the embodiment of the invention.Should understand can be by the flow process in each flow process in computer program instructions realization flow figure and/or the block scheme and/or square frame and process flow diagram and/or the block scheme and/or the combination of square frame.Can provide these computer program instructions to the processor of multi-purpose computer, special purpose computer, Embedded Processor or other programmable data processing device to produce a machine, make the instruction of carrying out through the processor of computing machine or other programmable data processing device produce to be used for the device of the function that is implemented in flow process of process flow diagram or a plurality of flow process and/or square frame of block scheme or a plurality of square frame appointments.

These computer program instructions also can be stored in ability vectoring computer or the computer-readable memory of other programmable data processing device with ad hoc fashion work; Make the instruction that is stored in this computer-readable memory produce the manufacture that comprises command device, this command device is implemented in the function of appointment in flow process of process flow diagram or a plurality of flow process and/or square frame of block scheme or a plurality of square frame.

These computer program instructions also can be loaded on computing machine or other programmable data processing device; Make on computing machine or other programmable devices and to carry out the sequence of operations step producing computer implemented processing, thereby the instruction of on computing machine or other programmable devices, carrying out is provided for being implemented in the step of the function of appointment in flow process of process flow diagram or a plurality of flow process and/or square frame of block scheme or a plurality of square frame.

Although described the preferred embodiments of the present invention, in a single day those skilled in the art get the basic inventive concept could of cicada, then can make other change and modification to these embodiment.So accompanying claims is intended to be interpreted as all changes and the modification that comprises preferred embodiment and fall into the scope of the invention.

Obviously, those skilled in the art can carry out various changes and modification to the present invention and not break away from the spirit and scope of the present invention.Like this, belong within the scope of claim of the present invention and equivalent technologies thereof if of the present invention these are revised with modification, then the present invention also is intended to comprise these changes and modification interior.

Claims (13)

1. a traffic information data restorative procedure is characterized in that, comprising:

Confirm traffic information data with the relevant road of target road correlativity maximum;

According to the regression coefficient between said target road and the said relevant road; And the current traffic information data of said relevant road; Confirm the traffic information data that said target road is current, said regression coefficient is confirmed with the historical traffic information data of said relevant road according to said target road.

2. the method for claim 1 is characterized in that, also comprises:

Confirm and the maximum relevant road of target road correlativity.

3. method as claimed in claim 2 is characterized in that, saidly confirms and the relevant road of target road correlativity maximum, specifically comprises:

The road of the setting quantity that temporal correlation definite and said target road is maximum, the tabulation of formation time correlativity, said temporal correlation is confirmed according to the speed time series of corresponding period of two road;

The road of the setting quantity that spatial coherence definite and said target road is maximum forms the spatial coherence tabulation, and said spatial coherence is confirmed according to the space length and the connectedness in two road;

Confirm and the maximum relevant road of target road correlation with said spatial coherence tabulation based on said temporal correlation tabulation.

4. method as claimed in claim 3 is characterized in that, the road of the setting quantity that said temporal correlation definite and said target road is maximum, and the tabulation of formation time correlativity specifically comprises:

To the road j identical in target road i and the road network, confirm that i, j two road speed seasonal effect in time series included angle cosine similarity coefficient are with target road i category of roads:

$s_{\mathrm{ij}}=\frac{\underset{t=1}{\overset{p}{\mathrm{\Σ}}}{x}_{\mathrm{it}}{x}_{\mathrm{jt}}}{\sqrt{\underset{t=1}{\overset{p}{\mathrm{\Σ}}}{x_{\mathrm{it}}}^2\·\underset{t=1}{\overset{p}{\mathrm{\Σ}}}{x_{\mathrm{jt}}}^2}}(i,j=1......n)$

Confirm that i, j two road speed seasonal effect in time series manhatton distance are:

$M_{\mathrm{dij}}=\underset{t=1}{\overset{p}{\mathrm{\Σ}}}|x_{\mathrm{it}}-x_{\mathrm{jt}}|(i,j=1......n)$

Settling the standard back i, j two road speed seasonal effect in time series standard manhatton distance are:

$R_{\mathrm{dij}}=\frac{\max \left(M_{\mathrm{dik}}\right)-M_{\mathrm{dij}}}{\max \left(M_{\mathrm{dik}}\right)-\min \left(M_{\mathrm{dik}}\right)}$

Confirm that i, j two road time correlation coefficient are:

C _ij＝ω·R _dij+θ·S _ij (ω+θ＝1，0≤ω≤1，0≤θ≤1)

To be recorded in the temporal correlation tabulation with the road of the maximum setting quantity of target road i time correlation coefficient;

Wherein, x _ItBe the speed time series of i bar road t each release cycle of period, x _JtBe the speed time series of road j to be determined at t each release cycle of period, i, j=1 ... N, n are the quantity of each grade road of system-wide net, and its value can obtain t=1 from the road network table ... P, p are the release cycle numbers of day part; M _DikBe other roads and the i bar road traffic flow speed seasonal effect in time series manhatton distance of the same link grade except j bar road in the road network, k=1 ... N, k ≠ i, k ≠ j; ω, θ are predefined weights.

5. method as claimed in claim 4 is characterized in that, the said t period is set according to traffic conditions and weather condition.

6. method as claimed in claim 3 is characterized in that, the road of the setting quantity that said spatial coherence definite and said target road is maximum forms the spatial coherence tabulation, specifically comprises:

To the road j identical in target road i and the road network with target road i category of roads, confirm the shortest operating range of road i and road j, confirm that the road of the setting quantity that the shortest said operating range is minimum is recorded in the spatial coherence tabulation.

7. method as claimed in claim 3 is characterized in that, and is said according to the relevant road that said temporal correlation tabulation is definite with said spatial coherence tabulation and the target road correlativity is maximum, specifically comprises:

Confirm the maximum a bar road of time correlation coefficient in the temporal correlation tabulation, confirm that the b bar road in the spatial coherence tabulation is relevant road in this a bar road, a is predefined value, b≤a;

When a and b are unequal; In other road in said spatial coherence tabulation except that said b bar road; Confirm the minimum road of the shortest operating range of a-b bar and said target road; The road travel permit number of determining when the tabulation based on said spatial coherence is during greater than a-b, in the road of selecting to determine based on said spatial coherence tabulation, and the a-b bar road that the time correlation coefficient is maximum;

In said b bar road and a-b bar road, confirm and the maximum relevant road of target road correlativity.

8. a traffic information data prosthetic device is characterized in that, comprising:

First data are confirmed the unit, are used for confirming the traffic information data with the relevant road of target road correlativity maximum;

Second data are confirmed the unit; Be used for according to the regression coefficient between said target road and the said relevant road; And the current traffic information data of said relevant road; Confirm the traffic information data that said target road is current, said regression coefficient is confirmed with the historical traffic information data of said relevant road according to said target road.

9. device as claimed in claim 8 is characterized in that, said first data confirm that the unit also is used for:

Confirm and the maximum relevant road of target road correlativity.

10. device as claimed in claim 9 is characterized in that, said first data confirm that the unit is confirmed and the relevant road of target road correlativity maximum, specifically comprise:

The road of the setting quantity that temporal correlation definite and said target road is maximum, the tabulation of formation time correlativity, said temporal correlation is confirmed according to the speed time series of corresponding period of two road;

The road of the setting quantity that spatial coherence definite and said target road is maximum forms the spatial coherence tabulation, and said spatial coherence is confirmed according to the space length and the connectedness in two road;

Confirm and the maximum relevant road of target road correlation with said spatial coherence tabulation based on said temporal correlation tabulation.

11. device as claimed in claim 10 is characterized in that, said first data confirm that the unit is confirmed and the road of the setting quantity that the temporal correlation of said target road is maximum, and the tabulation of formation time correlativity specifically comprises:

To the road j identical in target road i and the road network, confirm that i, j two road speed seasonal effect in time series included angle cosine similarity coefficient are with target road i category of roads:

$s_{\mathrm{ij}}=\frac{\underset{t=1}{\overset{p}{\mathrm{\Σ}}}{x}_{\mathrm{it}}{x}_{\mathrm{jt}}}{\sqrt{\underset{t=1}{\overset{p}{\mathrm{\Σ}}}{x_{\mathrm{it}}}^2\·\underset{t=1}{\overset{p}{\mathrm{\Σ}}}{x_{\mathrm{jt}}}^2}}(i,j=1......n)$

Confirm that i, j two road speed seasonal effect in time series manhatton distance are:

$M_{\mathrm{dij}}=\underset{t=1}{\overset{p}{\mathrm{\Σ}}}|x_{\mathrm{it}}-x_{\mathrm{jt}}|(i,j=1......n)$

Settling the standard back i, j two road speed seasonal effect in time series standard manhatton distance are:

$R_{\mathrm{dij}}=\frac{\max \left(M_{\mathrm{dik}}\right)-M_{\mathrm{dij}}}{\max \left(M_{\mathrm{dik}}\right)-\min \left(M_{\mathrm{dik}}\right)}$

Confirm that i, j two road time correlation coefficient are:

C _ij＝ω·R _dij+θ·S _ij (ω+θ＝1，0≤ω≤1，0≤θ≤1)

To be recorded in the temporal correlation tabulation with the road of the maximum setting quantity of target road i time correlation coefficient;

Wherein, x _ItBe the speed time series of i bar road t each release cycle of period, x _JtBe the speed time series of road j to be determined at t each release cycle of period, i, j=1 ... N, n are the quantity of each grade road of system-wide net, and its value can obtain t=1 from the road network table ... P, p are the release cycle numbers of day part; M _DikBe other roads and the i bar road traffic flow speed seasonal effect in time series manhatton distance of the same link grade except j bar road in the road network, k=1 ... N, k ≠ i, k ≠ j; ω, θ are predefined weights.

12. device as claimed in claim 10 is characterized in that, said first data confirm that the unit is confirmed and the road of the setting quantity that the spatial coherence of said target road is maximum, form the spatial coherence tabulation, specifically comprise:

To the road j identical in target road i and the road network with target road i category of roads, confirm the shortest operating range of road i and road j, confirm that the road of the setting quantity that the shortest said operating range is minimum is recorded in the spatial coherence tabulation.

13. device as claimed in claim 10 is characterized in that, said first data confirm that the unit is confirmed with said spatial coherence tabulation according to said temporal correlation tabulation and the relevant road of target road correlativity maximum, specifically comprise:

Confirm the maximum a bar road of time correlation coefficient in the temporal correlation tabulation, confirm that the b bar road in the spatial coherence tabulation is relevant road in this a bar road, a is predefined value, b≤a;

When a and b are unequal; In other road in said spatial coherence tabulation except that said b bar road; Confirm the minimum road of the shortest operating range of a-b bar and said target road; The road travel permit number of determining when the tabulation based on said spatial coherence is during greater than a-b, in the road of selecting to determine based on said spatial coherence tabulation, and the a-b bar road that the time correlation coefficient is maximum;

In said b bar road and a-b bar road, confirm and the maximum relevant road of target road correlativity.

Images