CN105139656A - Road state prediction method and device - Google Patents
Road state prediction method and device Download PDFInfo
- Publication number
- CN105139656A CN105139656A CN201510629427.2A CN201510629427A CN105139656A CN 105139656 A CN105139656 A CN 105139656A CN 201510629427 A CN201510629427 A CN 201510629427A CN 105139656 A CN105139656 A CN 105139656A
- Authority
- CN
- China
- Prior art keywords
- road
- predicted
- state
- value
- time
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Landscapes
- Traffic Control Systems (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The embodiments of the present invention disclose a road state prediction method and a device. The method comprises the steps of determining the state array of associated roads according to a to-be-predicted road, wherein the matrix value of the state array of associated roads is just the status value of roads that meet a preset distance relationship with the to-be-predicted road at a preset historical moment; calculating the to-be-predicted state of the to-be-predicted road according to the probabilistic relation between the to-be-predicted state of the to-be-predicted road at a to-be-predicted moment and the state array of associated roads. According to the technical scheme of the embodiments of the present invention, the state of the to-be-predicted road is predicted based on the probabilistic relation between the to-be-predicted state of the to-be-predicted road at the to-be-predicted moment and the state array of roads associated with the to-be-predicted road. Therefore, the road prediction accuracy is improved.
Description
Technical field
The embodiment of the present invention belongs to technical field of data processing, relates to a kind of road condition Forecasting Methodology and device.
Background technology
Accurately, the successful Application of telecommunication flow information to intelligent transportation system is most important timely.It can help road user to make better trip decision-making, alleviates traffic congestion, reduces carbon emission, and improves traffic circulation efficiency.Nowadays, traffic data becomes increasingly abundant, and we have entered the large data age of traffic.Effectively utilize the traffic flow forecasting that the large data of traffic are carried out more accurately and timely, supvr can be helped to make better traffic control scheme, and the trip decision-making for traffic trip person provides and provides powerful support for.
Existing traffic flow calculates and mainly calculates current load conditions, and object is can react current road condition comparatively accurately.Existing traffic flow forecasting method mainly utilizes the road conditions repeatability such as the periodicity of historical time framework to carry out the prediction of future period road conditions.
Existing traffic flow forecasting method only simply thinks that road conditions are repetitions of time, think that road conditions have time cycle property, too simple, Accurate Prediction road conditions are difficult under slightly complicated situation, not can solve the road condition predicting problem of actual complex, that is, at present good solution still be there is no for the prediction of following road conditions.
Summary of the invention
The object of the embodiment of the present invention proposes a kind of road condition Forecasting Methodology and device, to improve the accuracy of road condition prediction.
On the one hand, embodiments provide a kind of road condition Forecasting Methodology, comprising:
According to road determination associated road state array to be predicted, the matrix value of described associated road state array is the road meeting setpoint distance relation with described road to be predicted, and the state value within the setting historical juncture, described associated road state array is: S
m,n, wherein, the matrix value r of each matrix dot
i,jfor being the road of i with road distance to be predicted, the state value in j moment before current time;
Probabilistic relation between the state to be predicted of inscribing time to be predicted according to described road to be predicted and associated road state array, calculates described state to be predicted.
On the other hand, embodiments provide a kind of road condition prediction unit, comprising:
Association status array element, for according to road determination associated road state array to be predicted, the matrix value of described associated road state array is the road meeting setpoint distance relation with described road to be predicted, and the state value within the setting historical juncture, described associated road state array is: S
m,n, wherein, the matrix value r of each matrix dot
i,jfor being the road of i with road distance to be predicted, the state value in j moment before current time;
States prediction unit, for the probabilistic relation between the state to be predicted of inscribing time to be predicted according to described road to be predicted and associated road state array, calculates described state to be predicted.
The road condition Forecasting Methodology that the embodiment of the present invention provides and device, by the probabilistic relation between the state to be predicted of inscribing time to be predicted according to described road to be predicted and associated road state array, calculate the state to be predicted of road to be predicted, because the factor combining Time and place is predicted, therefore improve the accuracy of road condition prediction.
Accompanying drawing explanation
The schematic flow sheet of a kind of road condition Forecasting Methodology that Fig. 1 a provides for the embodiment of the present invention one;
A kind of road condition prediction that Fig. 1 b provides for the embodiment of the present invention one;
A kind of road condition prediction that Fig. 1 c provides for the embodiment of the present invention one;
The schematic flow sheet of a kind of road condition Forecasting Methodology that Fig. 2 provides for the embodiment of the present invention two;
The schematic flow sheet of a kind of road condition Forecasting Methodology that Fig. 3 a provides for the embodiment of the present invention three;
The road schematic diagram that Fig. 3 b provides for the embodiment of the present invention three;
The schematic flow sheet of a kind of road condition Forecasting Methodology that Fig. 4 provides for the embodiment of the present invention four;
The structural representation of a kind of road condition prediction unit that Fig. 5 provides for the embodiment of the present invention five.
Embodiment
Below in conjunction with drawings and Examples, the present invention is described in further detail.Be understandable that, specific embodiment described herein is only for explaining the present invention, but not limitation of the invention.It also should be noted that, for convenience of description, illustrate only part related to the present invention in accompanying drawing but not entire infrastructure.
Embodiment one
The schematic flow sheet of a kind of road condition Forecasting Methodology that Fig. 1 a provides for the embodiment of the present invention one.The present embodiment is applicable to user and wants predicting traffic flow or the situation by road conditions such as times.The method can be performed by road condition prediction unit.See Fig. 1 a, the road condition Forecasting Methodology that the present embodiment provides specifically comprises as follows:
S11, according to road determination associated road state array to be predicted, the matrix value of described associated road state array is the road meeting setpoint distance relation with described road to be predicted, the state value within the setting historical juncture.
Exemplary, described associated road state array can be: S
m,n, wherein, the matrix value r of each matrix dot
i,jfor being the road of i with road distance to be predicted, the state value in j moment before current time.M needs that set and between road to be predicted maximum distance according to predicting, n needs that set and between the moment to be predicted moment the earliest according to prediction, and i is less than or equal to m, and j is less than or equal to n.
In transportation network, road is interconnected.Composition graphs 1b and Fig. 1 c, the line of band arrow represents road and road direction, and circle represents the tie point between road, and wherein road is logical concept, and a logical road can comprise one or more physical road.Transportation network meets tie point flow Conservation Relationship, and namely tie point does not produce the magnitude of traffic flow, and the magnitude of traffic flow flowing into tie point equals the magnitude of traffic flow flowing out tie point, uses f
inrepresent the inflow flow of tie point, f
outrepresent the outflow flow of tie point, so following equation is set up:
f
in＝f
out
Every traffic flow situation of bar road is relevant with the traffic flow situation of the associated road of this road, in these linking-up roads, flow to this road traffic flow, also have the magnitude of traffic flow flowed out from this road, wherein, associated road can be the road having common connecting point with this road.The present embodiment utilizes magnitude of traffic flow conservation and road annexation to predict following road condition.
Composition graphs 1b and Fig. 1 c, the distance between the two road connected by a jumping is 2, and the distance between the two road connected by double bounce is 3, by that analogy, calculates the distance between road with jumping figure.The unit in moment can be setting value, such as 1 minute, 10 minutes etc.Road like state value can adopt the discrete value such as to block up, slow, unobstructed to represent, also can represent with average travelling speed value.
S12, probabilistic relation between the state to be predicted of inscribing time to be predicted according to described road to be predicted and associated road state array, calculate described state to be predicted.
Exemplary, the probabilistic relation between the state to be predicted of inscribing time to be predicted according to described road to be predicted and associated road state array, calculates described state to be predicted and specifically can comprise:
1) based on the expectation value of described probabilistic relation according to following formulae discovery state value to be predicted, as described state value to be predicted:
At state value consecutive hours, obtained the expectation value of state value to be predicted by integration according to above-mentioned formula, obtain the expectation value of state value to be predicted when state value is discrete according to discrete point summation.
2) according to described probabilistic relation, the state value corresponding to most probable value is chosen according to following formula, as state value to be predicted:
Wherein, P (R
k/ S
m,n) be state R to be predicted
kwith associated road state array S
m,nbetween probabilistic relation,
for state value to be predicted.Each moment road to be predicted and each state array have a probabilistic relation, and probability is expressed as curve or discrete data point, and transverse axis is the state value of moment road to be predicted, and the longitudinal axis is probable value; Probabilistic relation is determined by the statistics of historical data.
Exemplary, state value to be predicted can be traffic flow value to be predicted or time value of passing through to be predicted, namely, the traffic flow value of road to be predicted can be predicted by the traffic flow value of the association status battle array of road to be predicted, by the association status battle array of road to be predicted can predict road to be predicted by time value pass through time value.
The road condition Forecasting Methodology that the present embodiment provides, by according to road determination associated road state array to be predicted, and the probabilistic relation between the state to be predicted of inscribing time to be predicted according to described road to be predicted and associated road state array, calculate the state value of road to be predicted, because probabilistic relation has considered Time and place two dimensions, only consider that time cycle property improves the accuracy of road condition prediction compared to existing road Forecasting Methodology.
Embodiment two
The present embodiment provides a kind of new road condition Forecasting Methodology on the basis of above-described embodiment.The schematic flow sheet of a kind of road condition Forecasting Methodology that Fig. 2 provides for the embodiment of the present invention two.See Fig. 2, the road condition Forecasting Methodology that the present embodiment provides specifically comprises as follows:
S21, according to road determination associated road state array to be predicted, the matrix value of described associated road state array is the road meeting setpoint distance relation with described road to be predicted, the state value within the setting historical juncture.
On space structure, the state interact relation between road is different with the change of distance, and according to distance order from the close-by examples to those far off, the set expression being less than or equal to the road of m apart from the distance of a road l is L
m,
L
m＝{l
i},distance(l,l
i)≤m
In time structure, road l
ithe traffic flow modes of current time is l
i, 0, the traffic flow modes before the distance current time j moment is l
i,j.In Time and place two dimensions, distance is less than or equal to distance threshold values m, and the set that the time is not more than the traffic flow modes of the road of time threshold values n is S
m,n, be called traffic flow spatio-temporal state matrix, be called for short state array,
S
m,n＝{l
i,j},distance(l,l
i)≤mandj≤n
The associated road matrix S when threshold values n is different with the time for distance threshold values m
m,ndifference, that is, a road correspondence to be predicted at least two associated road matrixes,
with
represent any one state array at least two state array respectively.
S22, probabilistic relation between the state to be predicted of inscribing time to be predicted according to described road to be predicted and associated road state array, determine the state value to be predicted of associated road, and the state value to be predicted that will determine, as the predicted state value of associated road state array.
Concrete, according to the state R to be predicted that road to be predicted is inscribed time to be predicted
kwith associated road state array
between probabilistic relation
determine associated road state array
predicted state value, according to the state R to be predicted that road to be predicted is inscribed time to be predicted
kwith associated road state array
between probabilistic relation
determine associated road state array
predicted state value, that is, determine the predicted state value of each associated road state array.
S23, at least two associated road state array according to described road to be predicted, therefrom determine the associated road state array that matrix distance meets setting and requires.
Exemplary, at least two associated road state array of described road to be predicted can be: set road distance m with not meeting in the same time of described road to be predicted, set at least two associated road state array of historical juncture n.
Exemplary, the distance between state array adopts following formulae discovery:
with
represent any one state array at least two state array respectively,
I
m,n=(w
i,j), i ∈ [0, m], j ∈ [0, ∈ n], w
i,jrepresent road i apart from the traffic flow modes before the current j moment to the weighing factor of future transportation stream mode.
Exemplary, state value to be predicted is traffic flow value to be predicted, and in influence matrix, weighted value is determined according to the category of roads of road, road direction, road shape and/or site of road.Article one, the flow of the traffic conditions and it self in the future of road is relevant, also relevant with the traffic conditions of its other road related, affected degree is relevant to distance, larger apart from nearer impact, the impact of distance is relatively little, relation on this road space structure also determines traffic flow relation spatially, and the spatial relationship on this flow can be used for doing the volume forecasting of road.The magnitude of traffic flow of road is not only subject to the impact of road and traffic flow spatial relationship, also relevant to the time variations of road network traffic conditions, relation between the traffic flow modes that interval time is nearer is tightr, and the relation between the traffic flow modes that interval time is far away is relatively weak.This traffic flow modes also can be used for doing the volume forecasting of road in the correlationship of time.
The predicted state value corresponding to associated road state array that S24, acquisition are determined.
Obtain the predicting traffic flow value that matrix distance meets each associated road state array that setting requires.
S25, predicted state value according to the associated road state array determined, determine the state value to be predicted of described road to be predicted.
Concrete, KNN (k-NearestNeighbor can be adopted, contiguous) algorithm determines the traffic flow value to be predicted of described road to be predicted, the thinking of KNN algorithm chooses the maximum traffic flow modes of quantity corresponding to a nearest K state array, as the predicting traffic flow value of road to be predicted.Concrete, using the predicting traffic flow amount of predicting traffic flow amounts maximum for occurrence number in each predicting traffic flow value of determining in S24 as road to be predicted.
In addition, the mode additionally providing a kind of improvement determines the state value to be predicted of road to be predicted: after have chosen the adjacent state array of K, and with the traffic flow modes value of the mode computational prediction of linear weighted function, the vector that weighting coefficient is formed is called state composite vector.
In the present embodiment, influence matrix and state composite vector can adopt the method for machine learning to obtain, and do not limit to here and adopt which kind of machine learning method.Further, the exponent number m of space correlation and the exponent number n of time correlation can select according to actual needs, and the larger prediction of exponent number is more accurate, but complexity increases, and needs to weigh between this accuracy and complexity.
The road condition Forecasting Methodology that the present embodiment provides, the associated road state array of setting requirement is met by the distance of selection matrix from least two associated road state array, and determine the predicted state value of each associated road state array selected, determine the predicted state value of road to be predicted subsequently according to each predicted state value determined, improve the accuracy of road condition prediction.
Embodiment three
The present embodiment provides a kind of new road condition Forecasting Methodology on the basis of above-described embodiment, and state value to be predicted is to be predictedly pass through the time in the present embodiment.The schematic flow sheet of a kind of road condition Forecasting Methodology that Fig. 3 a provides for the embodiment of the present invention three.See Fig. 3 a, the road condition Forecasting Methodology that the present embodiment provides specifically comprises as follows:
S31, according to road determination associated road state array to be predicted, the matrix value of described associated road state array is the road meeting setpoint distance relation with described road to be predicted, within the setting historical juncture, pass through time value.
S32, according to described road to be predicted inscribe time to be predicted to be predicted by the probabilistic relation between time and associated road state array, calculate and describedly to be predictedly pass through the time.
Exemplary, can be comprised by time prediction for target road: described target road is carried out segmentation according to tie point; Piecewise prediction every section road by the time, predict using the terminal of road clearance time the last period as the initial time of next section of road, obtain the final time point arrived.Specifically also can be comprised by time prediction for target road: using described target road as road to be predicted, carry out whole section of prediction.
Exemplary, before prediction is by the time, can also comprise: the m choosing the associated road state array of road to be predicted according to the distance with road starting point, n value, wherein, the m that distance starting point is nearer, n value is less, the m that distance starting point is far away, and n value is larger.
Concrete, can there be two kinds of modes in path by time Estimate, segmentation estimation and overall estimation.As shown in Figure 3 b, the first half is that segmentation is estimated, the latter half is overall estimation.Segmentation method of estimation can carry out segmentation to the whole path of target road, different sections is estimated that it passes through the time, when making in this way, with initial time for zero, every period of moment that will estimate be all segmentations above by the time and, the parameter of each section can be selected separately.When using overall estimation, by the time, overall estimation is carried out to the whole path of target road.The selective dependency of Spatial dimensionality in the distance in prediction moment, large than path near-end of dimension needed for the transit time of path far-end, the size of univers parameter by near-end to far-end radially, design parameter select can adopt linearly radial or non-linear radial.
The road condition Forecasting Methodology that the present embodiment provides, by according to road determination associated road state array to be predicted, and the probabilistic relation between the state to be predicted of inscribing time to be predicted according to described road to be predicted and associated road state array, what calculate road to be predicted passes through time value, because probabilistic relation has considered Time and place two dimensions, only consider that time cycle property improves the accuracy of road condition prediction compared to existing road Forecasting Methodology.
Embodiment four
The present embodiment provides a kind of new road condition Forecasting Methodology on the basis of above-described embodiment three.The schematic flow sheet of a kind of road condition Forecasting Methodology that Fig. 4 provides for the embodiment of the present invention four.See Fig. 4, the road condition Forecasting Methodology that the present embodiment provides specifically comprises as follows:
S41, according to road determination associated road state array to be predicted, the matrix value of described associated road state array is the road meeting setpoint distance relation with described road to be predicted, within the setting historical juncture, pass through time value.
On space structure, the state interact relation between road is different with the change of distance, and according to distance order from the close-by examples to those far off, the set expression being less than or equal to the road of m apart from the distance of a road l is L
m,
L
m＝{l
i},distance(l,l
i)≤m
In time structure, road l
ithe traffic flow modes of current time is l
i, 0, the traffic flow modes before the distance current time j moment is l
i,j.In Time and place two dimensions, distance is less than or equal to distance threshold values m, and the set that the time is not more than the traffic flow modes of the road of time threshold values n is S
m,n, be called traffic flow spatio-temporal state matrix, be called for short state array,
S
m,n＝{l
i,j},distance(l,l
i)≤mandj≤n
, road correspondence to be predicted at least two associated road matrixes,
with
represent any one state array at least two state array respectively.
Under certain m and n, this road in a road following kth moment pass through time R
kp (R can be expressed as with probability with the Relation Between Traffic Flow of room and time distance within the scope of m and n
k/ S
m,n), be referred to as to pass through time probability.
S42, according to described road to be predicted inscribe time to be predicted to be predicted by the probabilistic relation between time and associated road state array, determine that the to be predicted of associated road passes through time value, and to be predicted by time value by what determine, time value is passed through in the prediction as associated road state array.
Exemplary, according to described road to be predicted inscribe time to be predicted to be predicted by the probabilistic relation between time and associated road state array, calculate described to be predictedly specifically can be comprised by the time:
1) based on described probabilistic relation according to the following formulae discovery expectation value by time value to be predicted, to be predictedly pass through time value as described:
2) according to described probabilistic relation, according to following formula choose corresponding to most probable value by time value, pass through time value as to be predicted:
Wherein, P (R
k/ S
m,n) be state R to be predicted
kwith associated road state array S
m,nbetween probabilistic relation,
for state value to be predicted.
S43, at least two associated road state array according to described road to be predicted, therefrom determine the associated road state array that matrix distance meets setting and requires.
Exemplary, at least two associated road state array of described road to be predicted can be: set road distance m with not meeting in the same time of described road to be predicted, set at least two associated road state array of historical juncture n.
Exemplary, the distance between state array adopts following formulae discovery:
with
represent any one state array at least two state array respectively,
I
m,n=(w
i,j), i ∈ [0, m], j ∈ [0, ∈ n], w
i,jrepresent road i apart from the traffic flow modes before the current j moment to the weighing factor of future transportation stream mode.
Time value is passed through in the prediction corresponding to associated road state array that S44, acquisition are determined.
S45, according to the prediction of the associated road state array determined by time value, determine that described the to be predicted of road to be predicted passes through time value.
S46, to drive correction factor according to individual to be predictedly being revised by the time of obtaining.
Road equals initial time and add upper pathway by the time time of arrival, and doping road clearance time namely can time of arrival of predicted link.Obtain collection routes by the time after, can be revised by the time set according to individual correction factor, and then obtain individual path by time or time of arrival, the preparation method of individual correction factor is not limited.
This time of arrival, method of estimation was based on space-time restriction relation, can according to different occasions, to the different demands of complexity and accuracy, adopted embodiment flexibly:
1) simply apply.Can build the model that typical road connects shape, such as linearly connected relational model, T-shaped link model, cruciform link model, annular link model, waits the link model of other real road.These Model Independents, in the position of road, when road clearance time is estimated, select the link model that road is corresponding.
2) slightly complicated application.Except road connects shape, by category of roads, the factors such as road direction are taken into account, and build the space-time restriction model connecting the composite factor such as shape, road type.This embodiment, also independent of the position of road, when road clearance time is estimated, selects the link model that road is corresponding.
3) what site of road was relevant implements in full mode.Space-time restriction model is built to every bar road, sets up the model bank that road is relevant, implement the personalized model towards road.This embodiment can the detail difference of accurate response diverse location road, reaches the effect accurately estimated.Method of estimation time of arrival that the present embodiment proposes can be applied to the various application that navigation, road conditions perception, share-car etc. arrive time correlation.
The technical scheme that the present embodiment provides, first estimates the rendezvous value of time of arrival, and the set obtaining path is estimated time of arrival, then adjusts according to the correction factor of individuality, obtains estimating individual time of arrival.Set time of arrival and the individual time of arrival in path are independently in application.The method has considered in transportation network, road is interconnected, the relation of the travel situations of vehicle on road both and between the driving habits of individual, have also contemplated that the relation between travel situations and the traffic of road, improves the accuracy of road clearance time prediction.
Embodiment five
The structural representation of the interworking unit of a kind of many system of account that Fig. 5 provides for the embodiment of the present invention five.The present embodiment is applicable to user and wants predicting traffic flow or the situation by road conditions such as times.See Fig. 5, the concrete structure of this road condition prediction unit is as follows:
Association status array element 51, for according to road determination associated road state array to be predicted, the matrix value of described associated road state array is the road meeting setpoint distance relation with described road to be predicted, the state value within the setting historical juncture;
States prediction unit 52, for the probabilistic relation between the state to be predicted of inscribing time to be predicted according to described road to be predicted and associated road state array, calculates described state to be predicted.
Exemplary, described associated road state array is:
S
m,n, wherein, the matrix value r of each matrix dot
i,jfor being the road of i with road distance to be predicted, the state value in j moment before current time.
Exemplary, described states prediction unit 52 specifically for:
Based on the expectation value of described probabilistic relation according to following formulae discovery state value to be predicted, as described state value to be predicted:
Or, according to described probabilistic relation, choose the state value corresponding to most probable value according to following formula, as state value to be predicted:
Wherein, P (R
k/ S
m,n) be state R to be predicted
kwith associated road state array S
m,nbetween probabilistic relation,
for state value to be predicted.
Exemplary, described states prediction unit 52 comprises:
Predicted state value subelement, for the state value to be predicted that will determine, as the predicted state value of associated road state array;
Association status frame subelement, at least two associated road state array according to described road to be predicted, therefrom determines that matrix distance meets the associated road state array of setting requirement;
Association status value subelement, for obtaining the predicted state value corresponding to the associated road state array determined;
Status predication subelement, for the predicted state value according to the associated road state array determined, determines the state value to be predicted of described road to be predicted.
Exemplary, at least two associated road state array of described road to be predicted are:
Set road distance m with not meeting in the same time of described road to be predicted, set at least two associated road state array of historical juncture n.
Exemplary, the distance between state array adopts following formulae discovery:
with
represent any one state array at least two state array respectively,
I
m,n=(w
i,j), i ∈ [0, m], j ∈ [0, ∈ n], w
i,jrepresent road i apart from the traffic flow modes before the current j moment to the weighing factor of future transportation stream mode.
Exemplary, in influence matrix, weighted value is determined according to the category of roads of road, road direction, road shape and/or site of road.
Exemplary, state value to be predicted is traffic flow value to be predicted or to be predictedly passes through time value.
Exemplary, when state value to be predicted be to be predicted by the time time, described states prediction unit for target road by time prediction specifically for:
Described target road is carried out segmentation according to tie point;
Piecewise prediction every section road by the time, predict using the terminal of road clearance time the last period as the initial time of next section of road, obtain the final time point arrived.
Exemplary, when state value to be predicted be to be predicted by the time time, described states prediction unit for target road by time prediction specifically for:
Using described target road as road to be predicted, carry out whole section of prediction.
Exemplary, this device also comprises:
Parameter choose unit, for before prediction is by the time, chooses the m of the associated road state array of road to be predicted, n value according to the distance with road starting point, wherein, the m that distance starting point is nearer, n value is less, the m that distance starting point is far away, and n value is larger.
Exemplary, drive correction factor to be predictedly being revised by the time of obtaining according to individual.
Above-mentioned road condition prediction unit can perform the road condition Forecasting Methodology that any embodiment of the present invention provides, and possesses the corresponding functional module of manner of execution and beneficial effect.The not ins and outs of detailed description in the present embodiment, the road condition Forecasting Methodology that can provide see any embodiment of the present invention.
Note, above are only preferred embodiment of the present invention and institute's application technology principle.Skilled person in the art will appreciate that and the invention is not restricted to specific embodiment described here, various obvious change can be carried out for a person skilled in the art, readjust and substitute and can not protection scope of the present invention be departed from.Therefore, although be described in further detail invention has been by above embodiment, the present invention is not limited only to above embodiment, when not departing from the present invention's design, can also comprise other Equivalent embodiments more, and scope of the present invention is determined by appended right.
Claims (22)
1. a road condition Forecasting Methodology, is characterized in that, comprising:
According to road determination associated road state array to be predicted, the matrix value of described associated road state array is the road meeting setpoint distance relation with described road to be predicted, and the state value within the setting historical juncture, described associated road state array is: S
m,n, wherein, the matrix value r of each matrix dot
i,jfor being the road of i with road distance to be predicted, the state value in j moment before current time;
Probabilistic relation between the state to be predicted of inscribing time to be predicted according to described road to be predicted and associated road state array, calculates described state to be predicted.
2. method according to claim 1, is characterized in that, the probabilistic relation between the state to be predicted of inscribing time to be predicted according to described road to be predicted and associated road state array, calculates the state of described moment road to be predicted, comprising:
Based on the expectation value of described probabilistic relation according to following formulae discovery state value to be predicted, as described state value to be predicted:
Or, according to described probabilistic relation, choose the state value corresponding to most probable value according to following formula, as state value to be predicted:
Wherein, P (R
k/
m,n) be state R to be predicted
kwith associated road state array S
m,nbetween probabilistic relation,
for state value to be predicted.
3. method according to claim 2, is characterized in that, the probabilistic relation between the state to be predicted of inscribing time to be predicted according to described road to be predicted and associated road state array, calculates described state to be predicted, comprising:
By the state value to be predicted determined, as the predicted state value of associated road state array;
According at least two associated road state array of described road to be predicted, therefrom determine that matrix distance meets the associated road state array of setting requirement;
Obtain the predicted state value corresponding to associated road state array determined;
According to the predicted state value of the associated road state array determined, determine the state value to be predicted of described road to be predicted.
4. method according to claim 3, is characterized in that, at least two associated road state array of described road to be predicted are:
Set road distance m with not meeting in the same time of described road to be predicted, set at least two associated road state array of historical juncture n.
5. method according to claim 3, is characterized in that, the distance between state array adopts following formulae discovery:
with
represent any one state array at least two state array respectively,
I
m,n=(w
i,j), i ∈ [0, m], j ∈ [0, ∈ n], w
i,jrepresent road i apart from the traffic flow modes before the current j moment to the weighing factor of future transportation stream mode.
6. method according to claim 5, is characterized in that, in influence matrix, weighted value is determined according to the category of roads of road, road direction, road shape and/or site of road.
7. method according to claim 1, is characterized in that, state value to be predicted is traffic flow value to be predicted or to be predictedly passes through time value.
8. method according to claim 7, is characterized in that, when state value to be predicted be to be predicted by the time time, being comprised by time prediction for target road:
Described target road is carried out segmentation according to tie point;
Piecewise prediction every section road by the time, predict using the terminal of road clearance time the last period as the initial time of next section of road, obtain the final time point arrived.
9. method according to claim 7, is characterized in that, when state value to be predicted be to be predicted by the time time, specifically being comprised by time prediction for target road:
Using described target road as road to be predicted, carry out whole section of prediction.
10. method according to claim 8 or claim 9, is characterized in that, before prediction is by the time, also comprises:
Choose the m of the associated road state array of road to be predicted according to the distance with road starting point, n value, wherein, the m that distance starting point is nearer, n value is less, the m that distance starting point is far away, and n value is larger.
11. methods according to claim 7, is characterized in that, drive correction factor to be predictedly being revised by the time of obtaining according to individual.
12. 1 kinds of road condition prediction units, is characterized in that, comprising:
Association status array element, for according to road determination associated road state array to be predicted, the matrix value of described associated road state array is the road meeting setpoint distance relation with described road to be predicted, and the state value within the setting historical juncture, described associated road state array is: S
m,n, wherein, the matrix value r of each matrix dot
i,jfor being the road of i with road distance to be predicted, the state value in j moment before current time;
States prediction unit, for the probabilistic relation between the state to be predicted of inscribing time to be predicted according to described road to be predicted and associated road state array, calculates described state to be predicted.
13. devices according to claim 12, is characterized in that, described states prediction unit specifically for:
Based on the expectation value of described probabilistic relation according to following formulae discovery state value to be predicted, as described state value to be predicted:
Or, according to described probabilistic relation, choose the state value corresponding to most probable value according to following formula, as state value to be predicted:
Wherein, P (R
k/ S
m,n) be state R to be predicted
kwith associated road state array S
m,nbetween probabilistic relation,
for state value to be predicted.
14. devices according to claim 13, is characterized in that, described states prediction unit comprises:
Predicted state value subelement, for the state value to be predicted that will determine, as the predicted state value of associated road state array;
Association status frame subelement, at least two associated road state array according to described road to be predicted, therefrom determines that matrix distance meets the associated road state array of setting requirement;
Association status value subelement, for obtaining the predicted state value corresponding to the associated road state array determined;
Status predication subelement, for the predicted state value according to the associated road state array determined, determines the state value to be predicted of described road to be predicted.
15. devices according to claim 14, is characterized in that, at least two associated road state array of described road to be predicted are:
Set road distance m with not meeting in the same time of described road to be predicted, set at least two associated road state array of historical juncture n.
16. devices according to claim 14, is characterized in that, the distance between state array adopts following formulae discovery:
with
represent any one state array at least two state array respectively,
I
m,n=(w
i,j), i ∈ [0, m], j ∈ [0, ∈ n], w
i,jrepresent road i apart from the traffic flow modes before the current j moment to the weighing factor of future transportation stream mode.
17. devices according to claim 16, is characterized in that, in influence matrix, weighted value is determined according to the category of roads of road, road direction, road shape and/or site of road.
18. devices according to claim 12, is characterized in that, state value to be predicted is traffic flow value to be predicted or to be predictedly passes through time value.
19. devices according to claim 18, is characterized in that, when state value to be predicted be to be predicted by the time time, described states prediction unit for target road by time prediction specifically for:
Described target road is carried out segmentation according to tie point;
Piecewise prediction every section road by the time, predict using the terminal of road clearance time the last period as the initial time of next section of road, obtain the final time point arrived.
20. devices according to claim 18, is characterized in that, when state value to be predicted be to be predicted by the time time, described states prediction unit for target road by time prediction specifically for:
Using described target road as road to be predicted, carry out whole section of prediction.
21. devices according to claim 19 or 20, is characterized in that, also comprise:
Parameter choose unit, for before prediction is by the time, chooses the m of the associated road state array of road to be predicted, n value according to the distance with road starting point, wherein, the m that distance starting point is nearer, n value is less, the m that distance starting point is far away, and n value is larger.
22. devices according to claim 18, is characterized in that, drive correction factor to be predictedly being revised by the time of obtaining according to individual.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510629427.2A CN105139656B (en) | 2015-09-28 | 2015-09-28 | A kind of road condition Forecasting Methodology and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510629427.2A CN105139656B (en) | 2015-09-28 | 2015-09-28 | A kind of road condition Forecasting Methodology and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105139656A true CN105139656A (en) | 2015-12-09 |
| CN105139656B CN105139656B (en) | 2017-09-29 |
Family
ID=54724983
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510629427.2A Active CN105139656B (en) | 2015-09-28 | 2015-09-28 | A kind of road condition Forecasting Methodology and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105139656B (en) |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106548625A (en) * | 2016-12-07 | 2017-03-29 | 山东易构软件技术股份有限公司 | A kind of urban highway traffic situation combination forecasting method |
| CN108986453A (en) * | 2018-06-15 | 2018-12-11 | 华南师范大学 | A kind of traffic movement prediction method based on contextual information, system and device |
| CN109300309A (en) * | 2018-10-29 | 2019-02-01 | 讯飞智元信息科技有限公司 | Road condition predicting method and device |
| CN109448381A (en) * | 2018-12-19 | 2019-03-08 | 安徽江淮汽车集团股份有限公司 | A kind of traffic prediction technique based on car networking big data |
| CN110751828A (en) * | 2019-09-10 | 2020-02-04 | 平安国际智慧城市科技股份有限公司 | Road congestion measuring method and device, computer equipment and storage medium |
| CN111091231A (en) * | 2019-11-25 | 2020-05-01 | 珠海格力电器股份有限公司 | Prediction model training method, time prediction method, training device and terminal |
| CN111653088A (en) * | 2020-04-21 | 2020-09-11 | 长安大学 | Vehicle driving quantity prediction model construction method, prediction method and system |
| WO2020191701A1 (en) * | 2019-03-28 | 2020-10-01 | 北京百度网讯科技有限公司 | Road conditions prediction method, apparatus, and device, and computer storage medium |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP1069405A2 (en) * | 1999-07-14 | 2001-01-17 | Kabushiki Kaisha Equos Research | Navigation method and system |
| DE10057796A1 (en) * | 2000-11-22 | 2002-05-23 | Daimler Chrysler Ag | Vehicle-specific dynamic traffic forecasting method by finding best-match load curve from historical load curves |
| CN101123038A (en) * | 2007-07-11 | 2008-02-13 | 山东省计算中心 | A dynamic information collection method for associated road segments of intersection |
| CN102110365A (en) * | 2009-12-28 | 2011-06-29 | 日电（中国）有限公司 | Road condition prediction method and road condition prediction system based on space-time relationship |
| CN104157139A (en) * | 2014-08-05 | 2014-11-19 | 中山大学 | Prediction method and visualization method of traffic jam |
| CN104408915A (en) * | 2014-11-05 | 2015-03-11 | 青岛海信网络科技股份有限公司 | Traffic state parameter estimation method and system |
| CN104809879A (en) * | 2015-05-14 | 2015-07-29 | 重庆大学 | Expressway road traffic state estimation method based on dynamic Bayesian network |
| CN104933858A (en) * | 2015-05-14 | 2015-09-23 | 浙江工业大学 | Space traffic characteristic Kernel-KNN matching road traffic state obtain method |
-
2015
- 2015-09-28 CN CN201510629427.2A patent/CN105139656B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP1069405A2 (en) * | 1999-07-14 | 2001-01-17 | Kabushiki Kaisha Equos Research | Navigation method and system |
| DE10057796A1 (en) * | 2000-11-22 | 2002-05-23 | Daimler Chrysler Ag | Vehicle-specific dynamic traffic forecasting method by finding best-match load curve from historical load curves |
| CN101123038A (en) * | 2007-07-11 | 2008-02-13 | 山东省计算中心 | A dynamic information collection method for associated road segments of intersection |
| CN102110365A (en) * | 2009-12-28 | 2011-06-29 | 日电（中国）有限公司 | Road condition prediction method and road condition prediction system based on space-time relationship |
| CN104157139A (en) * | 2014-08-05 | 2014-11-19 | 中山大学 | Prediction method and visualization method of traffic jam |
| CN104408915A (en) * | 2014-11-05 | 2015-03-11 | 青岛海信网络科技股份有限公司 | Traffic state parameter estimation method and system |
| CN104809879A (en) * | 2015-05-14 | 2015-07-29 | 重庆大学 | Expressway road traffic state estimation method based on dynamic Bayesian network |
| CN104933858A (en) * | 2015-05-14 | 2015-09-23 | 浙江工业大学 | Space traffic characteristic Kernel-KNN matching road traffic state obtain method |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106548625A (en) * | 2016-12-07 | 2017-03-29 | 山东易构软件技术股份有限公司 | A kind of urban highway traffic situation combination forecasting method |
| CN106548625B (en) * | 2016-12-07 | 2019-02-26 | 山东易构软件技术股份有限公司 | A kind of urban highway traffic situation combination forecasting method |
| CN108986453A (en) * | 2018-06-15 | 2018-12-11 | 华南师范大学 | A kind of traffic movement prediction method based on contextual information, system and device |
| CN109300309A (en) * | 2018-10-29 | 2019-02-01 | 讯飞智元信息科技有限公司 | Road condition predicting method and device |
| CN109448381A (en) * | 2018-12-19 | 2019-03-08 | 安徽江淮汽车集团股份有限公司 | A kind of traffic prediction technique based on car networking big data |
| WO2020191701A1 (en) * | 2019-03-28 | 2020-10-01 | 北京百度网讯科技有限公司 | Road conditions prediction method, apparatus, and device, and computer storage medium |
| US11823574B2 (en) | 2019-03-28 | 2023-11-21 | Beijing Baidu Netcom Science And Technology Co., Ltd. | Method and apparatus for prediction road condition, device and computer storage medium |
| CN110751828A (en) * | 2019-09-10 | 2020-02-04 | 平安国际智慧城市科技股份有限公司 | Road congestion measuring method and device, computer equipment and storage medium |
| CN111091231A (en) * | 2019-11-25 | 2020-05-01 | 珠海格力电器股份有限公司 | Prediction model training method, time prediction method, training device and terminal |
| CN111091231B (en) * | 2019-11-25 | 2022-04-15 | 珠海格力电器股份有限公司 | Prediction model training method, time prediction method, training device and terminal |
| CN111653088A (en) * | 2020-04-21 | 2020-09-11 | 长安大学 | Vehicle driving quantity prediction model construction method, prediction method and system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105139656B (en) | 2017-09-29 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105139656A (en) | Road state prediction method and device | |
| Jin et al. | A group-based traffic signal control with adaptive learning ability | |
| CN102110365B (en) | Road condition prediction method and road condition prediction system based on space-time relationship | |
| CN106207290B (en) | A kind of electric car charging aid decision optimization method based on multi-source data | |
| CN103077615B (en) | Online learning method for optimizing signalized intersection queuing length | |
| WO2019061933A1 (en) | Traffic signal chord panning control method and system | |
| EP2953110B1 (en) | Travel control device and travel control system | |
| CN102521989B (en) | Dynamic-data-driven highway-exit flow-quantity predicting method | |
| CN110414747B (en) | Space-time long-short-term urban pedestrian flow prediction method based on deep learning | |
| CN105513350A (en) | Time-phased multi-parameter short-term traffic flow prediction method based on time-space characteristics | |
| CN102867407B (en) | Multistep prediction method for effective parking space occupation rate of parking lot | |
| CN110570672B (en) | Regional traffic signal lamp control method based on graph neural network | |
| EP2187369A2 (en) | Traffic signals control system | |
| CN103903430A (en) | Dynamic fusion type travel time predicting method with multi-source and isomorphic data adopted | |
| CN112489426A (en) | Urban traffic flow space-time prediction scheme based on graph convolution neural network | |
| CN113516277B (en) | Internet intelligent traffic path planning method based on road network dynamic pricing | |
| CN101866143A (en) | Road traffic service level prediction method based on space-time characteristic aggregation | |
| CN103745602B (en) | A kind of traffic flow forecasting method average based on sliding window | |
| CN110428614A (en) | A kind of traffic congestion temperature spatio-temporal prediction method based on non-negative tensor resolution | |
| WO2021073716A1 (en) | Traffic reasoner | |
| CN107730893A (en) | A kind of shared bus website passenger flow forecasting for multidimensional characteristic of being gone on a journey based on passenger | |
| JP3509906B2 (en) | Road traffic signal control optimization device and road traffic flow control system using the same | |
| CN111524358B (en) | Regional radiation sexual communication flow prediction method | |
| Zheng et al. | A novel approach to coordinating green wave system with adaptation evolutionary strategy | |
| CN111341109A (en) | City-level signal recommendation system based on space-time similarity |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |