CN110517479A - A kind of urban highway traffic prediction technique, device and electronic equipment - Google Patents
A kind of urban highway traffic prediction technique, device and electronic equipment Download PDFInfo
- Publication number
- CN110517479A CN110517479A CN201810497797.9A CN201810497797A CN110517479A CN 110517479 A CN110517479 A CN 110517479A CN 201810497797 A CN201810497797 A CN 201810497797A CN 110517479 A CN110517479 A CN 110517479A
- Authority
- CN
- China
- Prior art keywords
- time
- dimension
- traffic data
- target road
- predicted
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/01—Detecting movement of traffic to be counted or controlled
- G08G1/0104—Measuring and analyzing of parameters relative to traffic conditions
- G08G1/0125—Traffic data processing
- G08G1/0129—Traffic data processing for creating historical data or processing based on historical data
Abstract
The embodiment of the present invention discloses a kind of urban highway traffic prediction technique, device and electronic equipment, is related to road traffic electric powder prediction, prediction result more accuracy.The method, comprising: obtain target road section urban road historical traffic flows data and speed data before current time；It is predicted according at least two different chronon dimensions and the historical traffic data, the traffic data to target road section in subsequent time, obtains the traffic data predicted value of time dimension；According to historical traffic data of the upstream section of target road section in current time previous predicted time interval, the traffic data to target road section in subsequent time is predicted, obtains the traffic data predicted value of Spatial Dimension；The weighted average for calculating the traffic data predicted value of the time dimension and the traffic data predicted value of the Spatial Dimension obtains target road section in the traffic data predicted value of subsequent time.The present invention is suitable for urban road short-time traffic flow forecast.
Description
Technical field
The present invention relates to road traffic electric powder prediction more particularly to a kind of urban highway traffic prediction techniques, device
And electronic equipment.
Background technique
Short-time traffic flow forecast research at present is than wide.Some is using time and spatiality mode point in the prior art
Not Huo get predicted value, predicted value is merged to obtain final predicted value, but this prediction technique only has chosen flow one
Traffic parameter is predicted, and is only to realize one-dimensional according to the historical standard sample database of foundation in the prediction of time dimension
The prediction of degree calculates, so that prediction result is not accurate enough.
Summary of the invention
In view of this, the embodiment of the present invention provides a kind of urban highway traffic prediction technique, device and electronic equipment, prediction
As a result more accurate.
In a first aspect, the embodiment of the present invention provides a kind of urban highway traffic prediction technique, comprising: obtain target road section and exist
Urban road historical traffic data before current time in predetermined time length；The historical traffic data is data on flows and speed
Degree evidence；According at least two different chronon dimensions and the historical traffic data, to target road section in subsequent time
Traffic data is predicted, the traffic data predicted value of time dimension is obtained；According to the upstream section of target road section when current
The historical traffic data in previous predicted time interval is carved, the traffic data to target road section in subsequent time is predicted, is obtained
Obtain the traffic data predicted value of Spatial Dimension；Calculate the traffic data predicted value of the time dimension and the friendship of the Spatial Dimension
The weighted average of logical data predicted value, obtains target road section in the traffic data predicted value of subsequent time.
A kind of specific implementation according to an embodiment of the present invention, at least two time dimension include timesharing dimension,
Its dimension and Zhou Weidu；It is described according to target road section current time at least two time dimensions forward historical traffic data,
Traffic data to target road section in subsequent time is predicted, comprising: according to target road section at current time forward at least two
The historical traffic data of a time dimension exists to target road section using the prediction model of the Zhou Weidu constructed based on k nearest neighbor algorithm
The traffic data of subsequent time is predicted.
A kind of specific implementation according to an embodiment of the present invention, the construction step of the prediction model of the Zhou Weidu, packet
Include: the urban road historical traffic data that will acquire, as unit of week carry out time series division, obtain as unit of week when
Between sequence；Determine the prediction more corresponding than number of days and timesharing dimension of the corresponding time series number of Zhou Weidu, the corresponding ring of day dimension
Time interval number；According to the time series number, current time corresponding speed state is obtained from the historical traffic data
Vector sum flow status vector；According to the time series number, the predicted time space-number and current time corresponding speed
State vector and flow status vector construct the Euclidean distance of the timesharing dimension of flow and speed, based on the ring than number of days and
The Euclidean distance of the day dimension of the Euclidean distance building flow and speed of the timesharing dimension, with the Euclidean distance of the day dimension
Inverse as weight, construct the flux prediction model of Zhou Weidu and the speed prediction model of Zhou Weidu respectively.
A kind of specific implementation according to an embodiment of the present invention, the upstream section according to target road section is when current
The historical traffic data in previous predicted time interval is carved, the traffic data to target road section in subsequent time is predicted, is obtained
Obtain the traffic data predicted value of Spatial Dimension, comprising: exist according to the section at least two layers of road network of target road section upstream neighbor
Historical traffic data in current time previous predicted time interval, the traffic data to target road section in subsequent time carry out in advance
It surveys, obtains the traffic data predicted value of Spatial Dimension.
A kind of specific implementation according to an embodiment of the present invention, at least two layers according to target road section upstream neighbor
Historical traffic data of the section in current time previous predicted time interval in road network, to target road section in subsequent time
Traffic data is predicted, the traffic data predicted value of Spatial Dimension is obtained, comprising: calculates upstream section and the mesh of target road section
Mark the Euclidean distance in section；The upstream section is the section at least two layers of road network of target road section upstream neighbor；According to upper
Swim actual flow of the section in the actual flow at current time and current time previous predicted time interval, with upstream section with
The inverse of the Euclidean distance of target road section calculates the traffic prediction value of the subsequent time of target road section as weight；According to upstream
Actual speed of the section in the actual speed at current time and current time previous predicted time interval, with upstream section and mesh
The inverse of the Euclidean distance in section is marked as weight, calculates the rate predictions of the subsequent time of target road section.
A kind of specific implementation according to an embodiment of the present invention, the traffic data prediction for calculating the time dimension
The weighted average of value and the traffic data predicted value of the Spatial Dimension, obtains target road section in the traffic data of subsequent time
Predicted value, comprising: calculate the flow related coefficient of time dimension and the flow related coefficient of Spatial Dimension；With the calculated time
The flow related coefficient of the flow related coefficient of dimension and Spatial Dimension as weight, calculate time dimension traffic prediction value and
The weighted average of the traffic prediction value of Spatial Dimension obtains target road section in the traffic prediction value of subsequent time；Calculate the time
The velocity correlation coefficint of dimension and the velocity correlation coefficint of Spatial Dimension；With the velocity correlation coefficint of calculated time dimension and
The velocity correlation coefficint of Spatial Dimension calculates the rate predictions of time dimension and the rate predictions of Spatial Dimension as weight
Weighted average, obtain target road section in the rate predictions of subsequent time.
A kind of specific implementation according to an embodiment of the present invention, the flow related coefficient and sky for calculating time dimension
Between dimension flow related coefficient, comprising: it is real according to the traffic prediction value of the time dimension in top n period at current time and flow
Actual value calculates the flow related coefficient of time dimension and the flow related coefficient of Spatial Dimension by Pearson correlation coefficient,
In, N is the natural number greater than 0；
The velocity correlation coefficint of the velocity correlation coefficint for calculating time dimension and Spatial Dimension, comprising: according to current
The rate predictions and speed actual value of the time dimension in top n period at moment calculate time dimension by Pearson correlation coefficient
The velocity correlation coefficint of degree and the velocity correlation coefficint of Spatial Dimension, wherein N is the natural number greater than 0.
Second aspect, the embodiment of the present invention provide a kind of urban highway traffic prediction meanss, comprising: traffic data obtains mould
Block, for obtaining urban road historical traffic data of the target road section before current time in predetermined time length；The history
Traffic data includes data on flows and speed data；Time dimension prediction module, for according at least two different chronons
Dimension and the historical traffic data, the traffic data to target road section in subsequent time is predicted, time dimension is obtained
Traffic data predicted value；Spatial Dimension prediction module, for the upstream section according to target road section in current time previous prediction
Historical traffic data in time interval, the traffic data to target road section in subsequent time are predicted, Spatial Dimension is obtained
Traffic data predicted value；Temporal-spatial fusion module, for calculate the time dimension traffic data predicted value and the space
The weighted average of the traffic data predicted value of dimension obtains target road section in the traffic data predicted value of subsequent time.
The third aspect, the embodiment of the present invention provide a kind of electronic equipment, and the electronic equipment includes: processor and storage
Device, wherein memory is for storing executable program code；Processor is by reading the executable program generation stored in memory
Code runs program corresponding with executable program code, for executing urban highway traffic described in aforementioned any implementation
Prediction technique.
Fourth aspect, the embodiment of the present invention also provide a kind of computer readable storage medium, described computer-readable to deposit
Storage media is stored with one or more program, and one or more of programs can be executed by one or more processor,
To realize urban highway traffic prediction technique described in aforementioned any claim.
A kind of urban highway traffic prediction technique, device and electronic equipment provided in an embodiment of the present invention, to target road
Section is when the time dimension of subsequent time and the traffic data predicted value of Spatial Dimension are predicted, due to the flow and speed of selection
Traffic parameters of the degree two with correlation, and the traffic data predicted value of time dimension be according at least two it is different when
Between the historical traffic data of sub- dimension carry out prediction acquisition so that according to the traffic data predicted value and sky of time dimension
Between the traffic data predicted value of dimension merged the target road section that obtains after (weighted average) in the traffic data of subsequent time
Predicted value is more accurate.
Detailed description of the invention
In order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, to embodiment or will show below
There is attached drawing needed in technical description to be briefly described, it should be apparent that, the accompanying drawings in the following description is only this
Some embodiments of invention for those of ordinary skill in the art without creative efforts, can be with
Other attached drawings are obtained according to these attached drawings.
Fig. 1 is the flow diagram of the urban highway traffic prediction technique of the embodiment of the present invention one；
Fig. 2 is the flow diagram of the urban highway traffic prediction technique of the embodiment of the present invention two；
Fig. 3 is the flow diagram for constructing prediction model in the embodiment of the present invention based on improved k nearest neighbor algorithm；
Fig. 4 is a urban road schematic diagram being made of three layers of road network in the embodiment of the present invention；
Fig. 5 is that the traffic data in the embodiment of the present invention on Spatial Dimension to target road section in subsequent time is predicted
Flow diagram；
Fig. 6 is that building process schematic diagram is gathered in middle and upper reaches of embodiment of the present invention section；
Fig. 7 is a urban road illustrative diagram in the embodiment of the present invention；
Fig. 8 is the urban highway traffic prediction meanss structural schematic diagram of the embodiment of the present invention three；
Fig. 9 is the urban highway traffic prediction meanss structural schematic diagram of the embodiment of the present invention four；
Figure 10 is the structural schematic diagram of electronic equipment one embodiment of the present invention.
Specific embodiment
The embodiment of the present invention is described in detail with reference to the accompanying drawing.
It will be appreciated that described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.Base
Embodiment in the present invention, it is obtained by those of ordinary skill in the art without making creative efforts it is all its
Its embodiment, shall fall within the protection scope of the present invention.
A kind of urban highway traffic prediction technique, device and electronic equipment provided in an embodiment of the present invention, applicable scene can
To be the prediction of short-term traffic volume of urban road.Traffic forecast refers in moment t to subsequent time t+1 or even several moment later
Real-time prediction is made in traffic flow.It is generally acknowledged that the predicted time span (i.e. predicted time interval) between t to t+1 is no more than 15 points
Clock is predicted as traffic parameter forecast in short-term.
Embodiment one
Fig. 1 is the flow diagram of the urban highway traffic prediction technique of the embodiment of the present invention one, as shown in Figure 1, this reality
The method for applying example may include:
Step 101 obtains urban road historical traffic data.
In the present embodiment, in order to realize the prediction of traffic data, need to obtain target road section pre- timing before current time
Between urban road historical traffic data in length.It is required to go through according to the length of predicted time for traffic forecast
The time span of history data has difference, and the present embodiment primarily directed to urban road, predicted in short-term by traffic, and acquisition is gone through
History data can be the historical traffic data of current time to the previous moon, two months, three months or five months, specific time span
Selection can select to determine according to actual needs.The historical traffic data of selection is flow quantity and speed data, the two friendships
Logical data can be obtained by the alert bayonet of electricity.Since speed and flow have correlation, predicted compared to independent using flow,
The present embodiment can more improve the accuracy of prediction.
Step 102, according at least two different chronon dimensions and historical traffic data, to target road section in lower a period of time
The traffic data at quarter is predicted, the traffic data predicted value of time dimension is obtained.
In the present embodiment, according to target road section current time forward at least two different chronon dimensions history hand over
Logical data, the traffic data to target road section in subsequent time is predicted, the traffic data predicted value of time dimension is obtained.
It specifically, can be according at least two not after obtaining the urban road historical traffic data in predetermined time length
The historical traffic data that same chronon dimension will acquire carries out division selection, and the history for obtaining at least two chronon dimensions is handed over
Logical data, with will pass through the historical traffic datas of at least two chronon dimensions to target road section subsequent time traffic data
It is predicted, obtains the traffic data predicted value on time dimension.In this way, predicted by the traffic data of multiple dimensions
As a result, it is possible to the traffic characteristics of more accurate response urban road, the accuracy of prediction and comprehensive is improved.
When the historical traffic data that will acquire divide according at least two different chronon dimensions and be chosen, can press
Timesharing dimension and day dimension carry out division selection, obtain the historical traffic data of two chronon dimensions, can also by timesharing dimension,
Its dimension and Zhou Weidu carry out division selection, obtain the historical traffic data of three chronon dimensions.
For by tri- timesharing dimension, day dimension and Zhou Weidu chronon dimensions, illustrate to historical traffic data
Specific division selection mode in temporal sequence: same day predetermined number (such as 6 is chosen forward using current time as initial time
It is a) predicted time interval (for example being 5 minutes) traffic data, obtain the same day timesharing dimension traffic data；With it is current when
Day where carving is the timesharing dimension traffic data for originating day and choosing preset number of days (such as 3 days) forward, obtains a day dimension traffic number
According to；The day dimension traffic data for choosing each week in predetermined time length for starting Zhou Xiangqian with week where current time, obtains week
Dimension traffic data.The selection of predetermined number, preset number of days, predetermined time period can select to determine according to actual needs.
Step 103, the historical traffic data according to the upstream section of target road section, to target road section subsequent time friendship
Logical data are predicted, the traffic data predicted value of Spatial Dimension is obtained.
In the present embodiment, handed over according to history of the upstream section of target road section in current time previous predicted time interval
Logical data, the traffic data to target road section in subsequent time is predicted, the traffic data predicted value of Spatial Dimension is obtained.
It specifically, can be according to the section in one or more layers road network of the upstream of target road section in current time previous prediction
Between historical traffic data in interval, the traffic data to target road section in subsequent time predicts, obtains in Spatial Dimension
On traffic data predicted value.
The weighting of the traffic data predicted value of step 104, the traffic data predicted value for calculating time dimension and Spatial Dimension
Average value obtains target road section in the traffic data predicted value of subsequent time.
In the present embodiment, by calculating the traffic data predicted value of the time dimension obtained and the traffic data of Spatial Dimension
Target road section can be obtained in the traffic data predicted value of subsequent time in the weighted average (i.e. temporal-spatial fusion) of predicted value.
The present embodiment urban highway traffic prediction technique, to target road section subsequent time time dimension and space dimension
When the traffic data predicted value of degree is predicted, due to the flow of selection and speed two traffic parameters with correlation, and
And the traffic data predicted value of time dimension is predicted according to the historical traffic data of the sub- dimension of at least two different times
It obtains, so that being merged according to the traffic data predicted value of the traffic data predicted value of time dimension and Spatial Dimension
The target road section obtained afterwards is more accurate in the traffic data predicted value of subsequent time.
Embodiment two
Fig. 2 is the flow diagram of the urban highway traffic prediction technique of the embodiment of the present invention two, as shown in Fig. 2, this reality
The method for applying example may include:
Step 201 obtains urban road historical traffic data.
In the present embodiment, the process of urban road historical traffic data and step 101 class of above method embodiment are obtained
Seemingly, details are not described herein again.
Step 202, according to the historical traffic data of at least two different chronon dimensions, to target road section in lower a period of time
The traffic data at quarter is predicted, the traffic data predicted value of time dimension is obtained.
In the present embodiment, according to target road section current time at least two chronon dimensions forward historical traffic number
According to the traffic data to target road section in subsequent time is predicted, the traffic data predicted value of time dimension is obtained.When passing through
Between the multiple sub- dimensions of dimension prediction, guarantee time dimension predicted value have accuracy and well adapting to property.
As an alternative embodiment, can be used improved k nearest neighbor algorithm to target road section subsequent time traffic data
It is predicted.In the present embodiment, the following aspects is mainly reflected in the improvement of k nearest neighbor algorithm:
(1) historical traffic data of acquisition is divided in temporal sequence, so that prediction result more preferably meets traffic flow
Moving law, to promote the science of prediction result；
(2) different weights are chosen and carries out various dimensions prediction, guarantee the accuracy of prediction；
(3) ring ratio was carried out to continuous X days (such as three days) data, prevented due to weather or unusual condition, caused pre-
The inaccuracy of result is surveyed, to promote the adaptability of prediction.
Wherein, the process historical traffic data of acquisition divided in temporal sequence, with drawing in above-described embodiment
Divide process identical, details are not described herein.
In the present embodiment, the traffic data using improved k nearest neighbor algorithm to target road section in subsequent time is predicted,
Specifically, be using the prediction model that is constructed based on improved k nearest neighbor algorithm, by the prediction model to target road section under
The traffic data at one moment is predicted.
Fig. 3 is the flow diagram that prediction model is constructed based on improved k nearest neighbor algorithm, referring to Fig. 3, is based on improved k
Nearest neighbor algorithm constructs prediction model, may include steps of:
S2021, the urban road historical traffic data that will acquire carry out time series division.
In the present embodiment, the urban road historical traffic data that will acquire is carried out time series division as unit of week, obtained
To the time series as unit of week.
Since urban road traffic flow characteristic was changed as unit of week, the time series of the present embodiment is drawn
Dividing is as unit of week, using the traffic parameter of each each period in week as an analytical unit.When encountering festivals or holidays, section is false
The previous day traffic stream characteristics that day has a holiday or vacation are according to the division of Friday, and workaday first day traffic stream characteristics are according to week after festivals or holidays
One division.
S2022, current time corresponding velocity state vectors and flow status vector are determined.
In order to provide data basis to the prediction of subsequent time, need to determine that current time is corresponding from historical traffic data
Velocity state vectors and flow status vector.
Wherein, what velocity state vectors indicated is the statistical value of current time corresponding speed, and statistical value here is pre-
Survey the average speed of time interval all vehicles such as in 5 minutes, such as: current time is 16:00, then speed at this time be from
15:55 assigns to average stroke speed of all vehicles on the section between 16:00.When what flow status vector indicated is current
The statistical value of flow is carved, traffic statistics value is the summation of all vehicle numbers in predicted time interval 5 minutes, such as: current time
It is 16:00, then flow at this time is to assign to flow summation of all vehicles on the section between 16:00 from 15:55.
It, can be according to Zhou Weidu in historical traffic data and the traffic flow data and number of speed of day dimension in the present embodiment
According to determine the velocity state vectors and flow status vector at current time.Specifically, it is handed in the urban road history that will acquire
Logical data, after carrying out time series division as unit of week, by weekly traffic flow data and speed data according to Monday
Sorted out to all object time sequences.The corresponding time series number of Zhou Weidu is determined, according to the corresponding time series of Zhou Weidu
Number obtains current time corresponding velocity state vectors and flow status vector from historical traffic data.Wherein, flow status
Vector can be denoted asVelocity state vectors can be denoted asWherein, t moment is current time, which in one week m be
It, h indicates the quantity (as time series number) in the week where current time, and this week h is 0, and h last week is 1.
Since the data that the present embodiment is selection 3 months are predicted that time series was divided as unit of week,
Therefore 3 months were counted according to 13 weeks in the present embodiment.
By being to construct prediction model based on improved k nearest neighbor algorithm in this present embodiment, the parameter in k nearest neighbor algorithm
K value and n value, the corresponding time series number of Zhou Weidu, the corresponding ring of day dimension corresponding to above-mentioned determination are than number of days and timesharing
The corresponding predicted time space-number of dimension, specifically, in the present embodiment, the meaning of k value and n value is as follows:
K is the number of selected forecast sample time series, i.e. the time series number of forecast sample, when corresponding to Zhou Weidu
Between sequence historical traffic data.What the present embodiment was chosen is 3 months data, and being assigned to time series is exactly 13 weeks, therefore k
Value be from 0 to 13.
N points are n1And n2Two classes.n1For the number at predicted time interval, i.e. predicted time space-number, correspond to timesharing dimension
The historical traffic data of time series.The n that the present embodiment is chosen1It is 6.If being divided into 5 minutes between predicted time, the present embodiment
What is chosen is the traffic data sample away from current time 30 minutes.n2For the number in the day of ring ratio, i.e. ring corresponds to day than number of days
The historical traffic data of dimension time series.The present embodiment, which is chosen, chose forward n away from the same day where current time t2A number of days
Traffic data carry out ring ratio, n in the present embodiment2Value is 4.
It should be understood that the value of above 3 parameters is only as exemplary illustration, it can also be according to the reality of different regions
Traffic condition carries out concrete configuration.
The Euclidean distance model of S2023, the multiple sub- dimensions of building.
The present embodiment contains three sub- dimensions in the prediction of time dimension, is k, the n mentioned in step S2022 respectively1、
n2Three sub- dimensions, k are the time series number of Zhou Weidu, n1For the predicted time space-number of timesharing dimension, n2For the ring of day dimension
Compare number of days.
When constructing the Euclidean distance model of multiple sub- dimensions, constructed from the smallest dimension of time series.
(1) the Euclidean distance model of timesharing dimension is constructed
Can according to the time series number, the predicted time space-number and current time corresponding velocity state vectors and
Flow status vector constructs the Euclidean distance of the timesharing dimension of flow and speed, specifically, can construct timesharing according to following formula
The Euclidean distance model of dimension:
Wherein,
To be all where current time, at a distance of the timesharing Euclidean distance of h weeks flow and speed；
H is the number in the week where current time, integer of the value between 0-13；
H is the mark of timesharing dimension；
I is the number at a distance of predicted time interval (such as 5 minutes) where current time, integer of the value between 0-5；
J is the number at a distance of day where current time, integer of the value between 0-3；
T is the timesharing time point where current time；
M is the day time point where current time；
λiFor at a distance of the timesharing weight of i-th of predicted time interval of current time (such as 5 minutes)；
qm-j(t-i) for apart from i-th of predicted time interval of current time (such as 5 minutes), the flow in jth day；
For apart from i-th of predicted time interval of current time (such as 5 minutes), jth day, the h weeks stream
Amount；
vm-j(T-i) for apart from i-th of predicted time interval of current time (such as 5 minutes), the speed in jth day；
For apart from i-th of predicted time interval of current time (such as 5 minutes), jth day, the h weeks stream
Amount；
For the flow average value of time dimension in i period；
For the speed average of time dimension in i period；
For the flow average value of the h weeks historical time dimension in i period；
For the speed average of the h weeks historical time dimension in i period；
n1Historical traffic data for the number at predicted time interval, corresponding to timesharing dimension time series.
In order to make parameterMeaning it is clearer, below with reference to h=2, when j=3, i=5 is selected to be gone through
History traffic data is to the parameter in above-mentioned formulaMeaning be illustrated:
If current time is 2018-05-0916:00, the flow status vector of current point in time is denoted as
When h=0, selected data are as shown in the table, and corresponding parameter and time meaning are as follows:
When h=1, apart from the data of current point in time the last week, selected data are as shown in the table, corresponding parameter and
Time meaning is as follows:
When h=2, data selected by the data apart from current point in time the last fortnight are as shown in the table, corresponding parameter and
Time meaning is as follows:
Similarly,Meaning it is same as above, only flow has changed speed into.
(2) the Euclidean distance model of day dimension is constructed
The Euclidean distance model of day dimension can be constructed according to following formula:
Wherein, DhIt (i) is the week where current time, at a distance of the day dimension Euclidean distance of h weeks flow and speed；
βjFor at a distance of the weight in current time jth day；
n2Historical traffic data for the number in the day of ring ratio, corresponding to day dimension time series；
The meaning of remaining symbol is same as above, and details are not described herein again.
The speed prediction model of S2024, the flux prediction model for constructing Zhou Weidu and Zhou Weidu.
In the present embodiment, using the inverse of the Euclidean distance of day dimension as weight, the volume forecasting of Zhou Weidu is constructed respectively
The speed prediction model of model and Zhou Weidu.
The present embodiment can construct the prediction model of Zhou Weidu, obtain by the inverse of the Euclidean distance of day dimension as weight
The predicted value of time dimension out.Specifically, the flux prediction model and Zhou Weidu of Zhou Weidu can be constructed respectively by following formula
Speed prediction model:
Wherein, QtimeIt (t+1) is subsequent time traffic flow forecasting value；
QhIt (t+1) is the week where current time, at a distance of h weeks subsequent time actual traffic amount；
VtimeIt (t+1) is subsequent time rate predictions；
VhIt (t+1) is the week where current time, at a distance of h weeks subsequent time actual speed；
K is the number of selected forecast sample time series, the historical traffic data corresponding to Zhou Weidu time series；
Remaining symbol meaning is same as above.
Step 203, according to the section at least two layers of road network of target road section upstream neighbor in current time previous prediction
Historical traffic data in time interval, the traffic data to target road section in subsequent time are predicted, Spatial Dimension is obtained
Traffic data predicted value.
Have section apart from short due to urban road, section branch is more, and DETECTION OF TRAFFIC PARAMETERS equipment is usually laid in road
Mouthful, therefore the upstream section that the present embodiment is chosen is made of n-layer road network.Fig. 4 is in the embodiment of the present invention one by three layers of road network
The urban road schematic diagram of composition.Referring to Fig. 4, there is shown with traversal point have first layer traversal point, second layer traversal point and the
Three layers of traversal point, the road shown have first layer road, second layer road and third layer road.
According to the laying point of urban highway traffic parameter acquisition device (electric police or flow detection device), if it is prediction
Time interval is 5 minutes, and the vehicle in upstream section was able to enter target road section within five minutes.
Fig. 5 is that the traffic data in the embodiment of the present invention on Spatial Dimension to target road section in subsequent time is predicted
Flow diagram.Referring to Fig. 5, as an alternative embodiment, according at least two layers of road network of target road section upstream neighbor
Historical traffic data of the section in current time previous predicted time interval, to target road section subsequent time traffic data
Predicted (step 203), it may include step:
S2031, building upstream section set.
Fig. 6 is that building process schematic diagram is gathered in middle and upper reaches of embodiment of the present invention section, referring to Fig. 6, constructs upstream section collection
The process of conjunction the following steps are included:
S20311, predicted time interval is obtained；
S20312, judge whether traversal point upstream adjacent segments have road network；
In this step, if traversal point upstream adjacent segments have road network, S20313 is thened follow the steps, upstream section is otherwise terminated
Traversal.
Speed of first section at current time in S20313, the adjacent road network in acquisition traversal point upstream；
S20314, the vehicle range ability traversed in the adjacent road network in point upstream in the first section is calculated；
S20315, judge the range ability whether less than the first section actual geographic length；
In this step, if the range ability is greater than or equal to the actual geographic length in the first section, show the first via
Vehicle in section can reach target road section in predicted time interval, and step S20316 thus can be performed and otherwise judge to traverse
Whether other sections in the adjacent road network in point upstream can be used as the element of upstream section set, until to the adjacent road in traversal point upstream
Each section judgement in net terminates.
S20316, the element for gathering in the first section as upstream section.
Wherein, the vehicle range ability in the adjacent road network in traversal point upstream in the first section can be calculated according to the following formula:
Wherein,
L(n1, n2..., nk) it is kth layer section, (n1, n2..., nk) a direction section vehicle range ability；
V(n1, n2..., nk) it is kth layer section, (n1, n2..., nk) a direction section running speed；
nkFor kth layer section, the number in n-th of direction.
Fig. 7 is a urban road illustrative diagram in the embodiment of the present invention.Below by taking road shown in Fig. 7 as an example, lift
Example illustrates the building process of upstream section set.
Referring to shown in Fig. 7, A1, A2, A3, A4 are four crossings, and L1, L2, L3 are three Duan Lu, and V1, V2, V3 are three sections of ways
Average speed is divided into T minutes between predicted time.
A1 is target crossing, and upstream section L1 is traversed by taking the crossing A1 as an example and is existed, determines the vehicle velocity V 1 of L1, calculates vehicle fortune
Whether row distance V1*T (i.e. V1 is multiplied by T) is greater than the geographical length of L1, if V1*T is greater than the geographical length of L1, using L1 as road
An element of Duan Jihe, and the upstream A2 section is found, if V1*T is less than the geographical length of L1, traverse this cut-off.
It is the geographical length that V1*T is greater than L1 in example, the element that L1 is gathered as section, and find the upstream A2 and have
Section L2 exists, and whether judgement (T-L1/V1) * V2 is greater than the geographical length of L2, if (T-L1/V1) * V2 is greater than the geography of L2
Length, another element that L2 is gathered as section, then the upstream A3 section L3 presence is found, judge (T-L1/V1-L2/V2) * V3
Whether L3 is greater than, if be less than L3, traversing third layer road network terminates, and finally obtaining section set is { L1, L2 }.
S2032, the Euclidean distance for calculating upstream section and target road section.
On the basis of the upstream section for generating target road section is gathered, disturbance degree of the upstream section to target road section is carried out
Analysis is calculated, the data of Y time interval of selected distance current time (such as 6 time intervals) are trained, by this Y
The target road section of time interval and its actual conditions in upstream section, calculate Euclidean distance, wherein Y is the nature greater than 0
Number.
As an alternative embodiment, the Euclidean distance of upstream section and target road section can be calculated according to the following formula:
Wherein,
DQV(n1, n2.., nk) be number be (n1, n2.., nk) flow in upstream section and the Euclidean distance of speed；
Q (t-i) is the actual flow of the target road section of i-th away from current time time interval；
qt-i-1(n1, n2.., nk) it is away from current time i+1 time interval, number is (n1, n2.., nk) upstream road
Section obtains actual flow；
V (t-i) is the actual speed of the target road section of i-th away from current time time interval；
For the flow average value of Spatial Dimension in i period；
For the speed average of Spatial Dimension in i period；
G is the quantity for selecting the period；
vt-i-1(n1, n2.., nk) it is away from current time i+1 time interval, number is (n1, n2.., nk) upstream road
Section obtains actual speed；
I is the number away from current time time interval；
λt-iFor the weight at t-i moment；
For the flow average value of the h weeks historical time dimension in i period；
For the speed average of the h weeks historical time dimension in i period.
S2033, the traffic parameter forecast value for calculating Spatial Dimension.
In the present embodiment, calculated according to the Euclidean distance that step S2032 is calculated using the inverse of Euclidean distance as weight
The predicted value of the subsequent time of target road section.
As an alternative embodiment, the traffic prediction value of the subsequent time of target road section can be calculated according to the following formula:
Wherein,
QspaceIt (t+1) is the predicted flow rate at t+1 moment；
qt(n1, n2.., nw) be sequence number be (n1, n2.., nw) section t moment actual flow；
qt-1(n1, n2.., nW) be sequence number be (n1, n2.., nW) the section t-1 moment actual flow；
M is the quantity of section set, and the value of M can be determined according to such as under type:
When the number of plies of road network is greater than 1,
When the number of plies of road network is equal to 1, M=4；
Parameter w is the number of plies of road network；
Remaining symbol is same as above.
As an alternative embodiment, the rate predictions of the subsequent time of target road section can be calculated according to the following formula:
Wherein, VspaceIt (t+1) is the predetermined speed at t+1 moment；
vt(n1, n2.., nw) be sequence number be (n1, n2.., nw) section t moment actual speed；
vt-1(n1, n2.., nw) be sequence number be (n1, n2.., nw) the section t-1 moment actual speed；
M is the quantity of element in the set of section, and the value of M can be determined according to such as under type:
When the number of plies of road network is greater than 1,
When the number of plies of road network is equal to 1, M=4:
Parameter w is the number of plies of road network；
Remaining symbol is same as above.
The traffic data prediction of step 204, the traffic data predicted value for calculating the time dimension and the Spatial Dimension
The weighted average of value obtains target road section in the traffic data predicted value of subsequent time.
As an alternative embodiment, traffic data predicted value and the space of time dimension can be calculated by temporal-spatial fusion algorithm
The weighted average of the traffic data predicted value of dimension obtains target road section in the traffic data predicted value of subsequent time, specifically
For, temporal-spatial fusion algorithm is the weight that time dimension and Spatial Dimension predicted value are calculated by Pearson correlation coefficient, then
Multiplied by the predicted value of corresponding t+1 time dimension and Spatial Dimension, the predicted value that the t+1 moment merges is finally obtained.
Optionally, the traffic data of the traffic data predicted value for calculating the time dimension and the Spatial Dimension is pre-
The weighted average of measured value obtains target road section in the traffic data predicted value (step 204) of subsequent time, it may include:
The flow related coefficient of S2041a, the flow related coefficient for calculating time dimension and Spatial Dimension.
As an alternative embodiment, the flow phase relation of the flow related coefficient and Spatial Dimension for calculating time dimension
Number, it may include: according to the traffic prediction value and flow actual value of the time dimension in top n period at current time, pass through Pearson came
Related coefficient calculates the flow related coefficient of time dimension and the flow related coefficient of Spatial Dimension, wherein N is oneself greater than 0
So number.
Optionally, the flow phase relation of the flow related coefficient and Spatial Dimension of time dimension can be calculated according to the following formula
Number:
Wherein,
For the flow related coefficient of time dimension；
For the flow related coefficient of Spatial Dimension；
For the time dimension apart from i period (i.e. i-th of time interval before current time) of current time
Actual flow, for example, the actual flow of the time dimension at current time isThe time dimension of previous time interval
Actual flow beThe actual flow of the time dimension of the first two time interval is
For the actual mean flow rate of the time dimension in i period；
For the predicted flow rate apart from the time dimension in i period of current time；
For the average flow rate of the time dimension prediction in i period；
For the actual flow apart from the Spatial Dimension in i period of current time；
For the actual mean flow rate of the Spatial Dimension in i period；
For the predicted flow rate apart from the Spatial Dimension in i period of current time；
For the prediction average flow rate of the Spatial Dimension in i period；
I is the quantity at the predicted time interval chosen；
R is the quantity for selecting time interval, such as when the quantity of 6 time intervals being taken to be predicted, the value of r=6, i
It is incremented by from 0 to 5.
S2042a, using the flow related coefficient of the flow related coefficient of calculated time dimension and Spatial Dimension as power
Weight, calculates the weighted average of the traffic prediction value of time dimension and the traffic prediction value of Spatial Dimension, obtains target road section and exist
The traffic prediction value of subsequent time.
As an alternative embodiment, the traffic prediction value and Spatial Dimension of time dimension can according to the following formula, be calculated
The weighted average of traffic prediction value obtains target road section in the traffic prediction value of subsequent time:
Wherein,For the t+1 moment, predicted flow rate after temporal-spatial fusion, i.e., target road section is in subsequent time
Traffic prediction value.
One t+1 moment, predetermined speed after temporal-spatial fusion；
R is the quantity for selecting the period；
Remaining symbol is same as above.
Optionally, the traffic data of the traffic data predicted value for calculating the time dimension and the Spatial Dimension is pre-
The weighted average of measured value obtains target road section in the traffic data predicted value (step 204) of subsequent time, it may include:
The velocity correlation coefficint of S2041b, the velocity correlation coefficint for calculating time dimension and Spatial Dimension.
As an alternative embodiment, the speed phase relation of the velocity correlation coefficint and Spatial Dimension for calculating time dimension
Number, it may include: according to the rate predictions and speed actual value of the time dimension in top n period at current time, pass through Pearson came
Related coefficient calculates the velocity correlation coefficint of time dimension and the velocity correlation coefficint of Spatial Dimension.
Optionally, the speed phase relation of the velocity correlation coefficint and Spatial Dimension of time dimension can be calculated according to the following formula
Number:
Wherein,
For time dimension velocity correlation coefficint；
For Spatial Dimension velocity correlation coefficint；
For apart from the time dimension actual speed in i period of current time；
For the time dimension actual average speed in i period；
For apart from the time dimension predetermined speed in i period of current time；
For the time dimension predicted average speed in i period；
For apart from the Spatial Dimension actual speed in i period of current time；
For the Spatial Dimension actual average speed in i period；
For apart from the Spatial Dimension predetermined speed in i period of current time；
For the Spatial Dimension predicted average speed in i period；
I is the quantity for the time interval chosen；
R is the quantity for selecting time interval, such as when the quantity of 6 time intervals being taken to be predicted, the value of r=6, i
It is incremented by from 0 to 5.
S2042b, using the velocity correlation coefficint of the velocity correlation coefficint of calculated time dimension and Spatial Dimension as power
Weight, calculates the weighted average of the rate predictions of time dimension and the rate predictions of Spatial Dimension, obtains target road section and exist
The rate predictions of subsequent time.
As an alternative embodiment, the rate predictions and Spatial Dimension of time dimension can according to the following formula, be calculated
The weighted average of rate predictions obtains target road section in the rate predictions of subsequent time:
Wherein,For the t+1 moment, predetermined speed after temporal-spatial fusion, i.e., target road section is in subsequent time
Rate predictions；
R is the quantity for selecting the period.
Remaining symbol is same as above.
The present embodiment urban highway traffic prediction technique, to target road section subsequent time time dimension and space dimension
When the traffic data predicted value of degree is predicted, due to the flow of selection and speed two traffic parameters with correlation, and
And the traffic data predicted value of time dimension is the history according to target road section in current time at least two time dimensions forward
Traffic data carries out prediction acquisition, so that according to the traffic number of the traffic data predicted value of time dimension and Spatial Dimension
It is predicted that the target road section that value obtains after being merged is more accurate in the traffic data predicted value of subsequent time.
In addition to this, the embodiment of the present invention also has the following beneficial effects:
Promote the adaptability of prediction result.During time dimension modeling, divided week, day, timesharing three it is different
Sub- dimension superposition is calculated, and in Spatial Dimension modeling process, is successively traversed, is combined to the upstream section of target road section
The case where urban highway traffic parameter detection equipment is laid, so that prediction result can adapt to the reasons such as complicated road and environment
It is influenced caused by prediction result.
The reliability of prediction result is improved, the embodiment of the present invention carries out pre- from two dimensions of time dimension and Spatial Dimension
It surveys, avoids the one-sidedness generated due to single dimension prediction result.
Embodiment three
Fig. 8 is the urban highway traffic prediction meanss structural schematic diagram of the embodiment of the present invention three, as shown in figure 8, this implementation
Example device may include: traffic data obtain module 11, time dimension prediction module 12, Spatial Dimension prediction module 13 and when
Empty Fusion Module 14.
Traffic data obtains module 11, for obtaining city road of the target road section before current time in predetermined time length
Road historical traffic data；The historical traffic data includes data on flows and speed data；
Time dimension prediction module 12, for according at least two different chronon dimensions and the historical traffic number
According to the traffic data to target road section in subsequent time is predicted, the traffic data predicted value of time dimension is obtained；
Spatial Dimension prediction module 13, for the upstream section according to target road section between current time previous predicted time
Every interior historical traffic data, the traffic data to target road section in subsequent time is predicted, obtains the traffic of Spatial Dimension
Data predicted value；
Temporal-spatial fusion module 14, for calculating the traffic data predicted value of the time dimension and the friendship of the Spatial Dimension
The weighted average of logical data predicted value, obtains target road section in the traffic data predicted value of subsequent time.
The device of the present embodiment can be used for executing the technical solution of embodiment of the method shown in Fig. 1, realization principle and skill
Art effect is similar, and details are not described herein again.
Example IV
Fig. 9 is the urban highway traffic prediction meanss structural schematic diagram of the embodiment of the present invention four, as shown in figure 9, this implementation
On the basis of three described device of embodiment, in the present embodiment, the time dimension prediction module 12 is specifically used for the device of example
It is right using the prediction model constructed by k nearest neighbor algorithm according to the historical traffic data of at least two different chronon dimensions
Target road section is predicted in the traffic data of subsequent time, obtains subsequent time traffic flow forecasting value and subsequent time speed
Predicted value.
As an alternative embodiment, at least two time dimension includes timesharing dimension, day dimension and Zhou Weidu；It is described
Time dimension prediction module 12, specifically for according to target road section current time forward at least two time dimensions history hand over
Logical data, using the prediction model of the Zhou Weidu constructed based on k nearest neighbor algorithm, to target road section subsequent time traffic data
It is predicted.
As an alternative embodiment, the urban highway traffic prediction meanss, further includes: prediction model constructs module
15, for constructing the prediction model of Zhou Weidu: the urban road historical traffic data that will acquire according to following steps, with Zhou Weidan
Position carries out time series division, obtains the time series as unit of week；Determine the corresponding time series number of Zhou Weidu, day dimension
Corresponding ring predicted time space-number more corresponding than number of days and timesharing dimension；According to the time series number, from the history
Current time corresponding velocity state vectors and flow status vector are obtained in traffic data；According to the time series number, institute
Predicted time space-number and current time corresponding velocity state vectors and flow status vector are stated, point of flow and speed is constructed
The Euclidean distance of Shi Weidu, the day dimension based on the ring than number of days and the Euclidean distance of timesharing dimension building flow and speed
The Euclidean distance of degree；Zhou Wei is constructed respectively using the inverse of the Euclidean distance of the day dimension as weight based on k nearest neighbor algorithm
The flux prediction model of degree and the speed prediction model of Zhou Weidu.
As an alternative embodiment, the Spatial Dimension prediction module 13 is specifically used for according to target road section upstream neighbor
At least two layers of road network in historical traffic data of the section in current time previous predicted time interval, exist to target road section
The traffic data of subsequent time is predicted, the traffic data predicted value of Spatial Dimension is obtained.
As an alternative embodiment, the Spatial Dimension prediction module 13, including Euclidean distance computational submodule 131 and stream
Amount prediction submodule 132；Wherein, Euclidean distance computational submodule 131, for calculating the upstream section and target road of target road section
The Euclidean distance of section；The upstream section is the section at least two layers of road network of target road section upstream neighbor；Volume forecasting
Module 132, for the reality according to upstream section in the actual flow at current time and current time previous predicted time interval
Border flow calculates the subsequent time of target road section using the inverse of upstream section and the Euclidean distance of target road section as weight
Traffic prediction value；
Prediction of speed submodule 133, for previous in the actual speed at current time and current time according to upstream section
Actual speed in predicted time interval calculates mesh using the inverse of upstream section and the Euclidean distance of target road section as weight
Mark the rate predictions of the subsequent time in section.
As an alternative embodiment, the temporal-spatial fusion module 14, including the first related coefficient computational submodule 141 and
One fusion submodule 142；Wherein, the first related coefficient computational submodule 141, for calculating the flow related coefficient of time dimension
With the flow related coefficient of Spatial Dimension；First fusion submodule 142, when being used for calculated with related coefficient computational submodule
Between dimension flow related coefficient and Spatial Dimension flow related coefficient as weight, calculate the traffic prediction value of time dimension
With the weighted average of the traffic prediction value of Spatial Dimension, target road section is obtained in the traffic prediction value of subsequent time.
As an alternative embodiment, the first related coefficient computational submodule 141, before being specifically used for according to current time
The traffic prediction value and flow actual value of the time dimension in N number of period calculate the stream of time dimension by Pearson correlation coefficient
Measure the flow related coefficient of related coefficient and Spatial Dimension.
As an alternative embodiment, the temporal-spatial fusion module 14, including the second related coefficient computational submodule 143 and
Two fusion submodules 144；Wherein, the second related coefficient computational submodule 143, for calculating the velocity correlation coefficint of time dimension
With the velocity correlation coefficint of Spatial Dimension；Second fusion submodule 144, for the speed phase relation of calculated time dimension
For several and Spatial Dimension velocity correlation coefficint as weight, the speed of the rate predictions and Spatial Dimension that calculate time dimension is pre-
The weighted average of measured value obtains target road section in the rate predictions of subsequent time.
As an alternative embodiment, the second related coefficient computational submodule 143, before being specifically used for according to current time
The rate predictions and speed actual value of the time dimension in N number of period calculate the speed of time dimension by Pearson correlation coefficient
Spend the velocity correlation coefficint of related coefficient and Spatial Dimension.
The device of the present embodiment can be used for executing the technical solution of embodiment of the method shown in Fig. 1 or Fig. 2, realize former
Reason is similar with technical effect, and details are not described herein again.
The embodiment of the present invention also provides a kind of electronic equipment, and the electronic equipment includes dress described in aforementioned any embodiment
It sets.
Figure 10 is the structural schematic diagram of electronic equipment one embodiment of the present invention, be may be implemented shown in Fig. 1 and Fig. 2 of the present invention
The process of embodiment, as shown in Figure 10, above-mentioned electronic equipment may include: shell 41, processor 42, memory 43, circuit board
44 and power circuit 45, wherein circuit board 44 is placed in the space interior that shell 41 surrounds, and processor 42 and memory 43 are arranged
On circuit board 44；Power circuit 45, for each circuit or the device power supply for above-mentioned electronic equipment；Memory 43 is for depositing
Store up executable program code；Processor 42 runs and can be performed by reading the executable program code stored in memory 43
The corresponding program of program code, for executing prediction technique described in aforementioned any embodiment.
Processor 42 to the specific implementation procedures of above-mentioned steps and processor 42 by operation executable program code come
The step of further executing may refer to the description of Fig. 1-2 illustrated embodiment of the present invention, and details are not described herein.
The electronic equipment exists in a variety of forms, including but not limited to:
(1) mobile communication equipment: the characteristics of this kind of equipment is that have mobile communication function, and to provide speech, data
Communication is main target.This Terminal Type includes: smart phone (such as iPhone), multimedia handset, functional mobile phone and low
Hold mobile phone etc..
(2) PC device: this kind of equipment has calculating and processing function, generally also has mobile Internet access characteristic.This
Terminal Type includes: PDA, MID and UMPC equipment etc., such as iPad.
(3) server: providing the equipment of the service of calculating, and the composition of server includes that processor, hard disk, memory, system are total
Line etc., server is similar with general computer architecture, but due to needing to provide highly reliable service, in processing energy
Power, stability, reliability, safety, scalability, manageability etc. are more demanding.
(4) other electronic equipments with data interaction function.
It should be noted that, in this document, relational terms such as first and second and the like are used merely to a reality
Body or operation are distinguished with another entity or operation, are deposited without necessarily requiring or implying between these entities or operation
In any actual relationship or order or sequence.Moreover, the terms "include", "comprise" or its any other variant are intended to
Non-exclusive inclusion, so that the process, method, article or equipment including a series of elements is not only wanted including those
Element, but also including other elements that are not explicitly listed, or further include for this process, method, article or equipment
Intrinsic element.In the absence of more restrictions, the element limited by sentence " including one ... ", it is not excluded that
There is also other identical elements in the process, method, article or apparatus that includes the element.
Each embodiment in this specification is all made of relevant mode and describes, same and similar portion between each embodiment
Dividing may refer to each other, and each embodiment focuses on the differences from other embodiments.
For Installation practice, since it is substantially similar to the method embodiment, so the comparison of description is simple
Single, the relevent part can refer to the partial explaination of embodiments of method.
For convenience of description, description apparatus above is to be divided into various units/modules with function to describe respectively.Certainly, In
Implement to realize each unit/module function in the same or multiple software and or hardware when the present invention.
Those of ordinary skill in the art will appreciate that realizing all or part of the process in above-described embodiment method, being can be with
Relevant hardware is instructed to complete by computer program, the program can be stored in a computer-readable storage medium
In, the program is when being executed, it may include such as the process of the embodiment of above-mentioned each method.Wherein, the storage medium can be magnetic
Dish, CD, read-only memory (Read-Only Memory, ROM) or random access memory (Random Access
Memory, RAM) etc..
The above description is merely a specific embodiment, but scope of protection of the present invention is not limited thereto, any
In the technical scope disclosed by the present invention, any changes or substitutions that can be easily thought of by those familiar with the art, all answers
It is included within the scope of the present invention.Therefore, protection scope of the present invention should be subject to the protection scope in claims.
Claims (10)
1. a kind of urban highway traffic prediction technique characterized by comprising
Obtain urban road historical traffic data of the target road section before current time in predetermined time length；The historical traffic
Data are data on flows and speed data；
According at least two different chronon dimensions and the historical traffic data, to target road section subsequent time traffic
Data are predicted, the traffic data predicted value of time dimension is obtained；
According to historical traffic data of the upstream section of target road section in current time previous predicted time interval, to target road
Section is predicted in the traffic data of subsequent time, obtains the traffic data predicted value of Spatial Dimension；
Calculate the weighted average of the traffic data predicted value of the time dimension and the traffic data predicted value of the Spatial Dimension
Value, obtains target road section in the traffic data predicted value of subsequent time.
2. urban highway traffic prediction technique according to claim 1, which is characterized in that described at least two it is different when
Between sub- dimension include timesharing dimension, day dimension and Zhou Weidu；
It is described according to the sub- dimension of at least two different times and the historical traffic data, to target road section subsequent time friendship
Logical data are predicted, comprising:
According to the sub- dimension of at least two different times and the historical traffic data, using the Zhou Wei constructed based on k nearest neighbor algorithm
The prediction model of degree, the traffic data to target road section in subsequent time are predicted.
3. urban highway traffic prediction technique according to claim 2, which is characterized in that the prediction model of the Zhou Weidu
Construction step, comprising:
The urban road historical traffic data that will acquire is carried out time series division as unit of week, obtained as unit of week
Time series；
Determine the corresponding time series number of Zhou Weidu, the corresponding ring of day dimension predicted time more corresponding than number of days and timesharing dimension
Space-number；
According to the time series number, current time corresponding velocity state vectors and stream are obtained from the historical traffic data
Measure state vector；
According to the time series number, the predicted time space-number and current time corresponding velocity state vectors and flow shape
State vector constructs the Euclidean distance of the timesharing dimension of flow and speed, the Europe based on the ring than number of days and the timesharing dimension
The Euclidean distance of the day dimension of formula distance building flow and speed, using the inverse of the Euclidean distance of the day dimension as weight,
The flux prediction model of Zhou Weidu and the speed prediction model of Zhou Weidu are constructed respectively.
4. urban highway traffic prediction technique according to claim 1, which is characterized in that described according to the upper of target road section
Swim historical traffic data of the section in current time previous predicted time interval, to target road section subsequent time traffic number
According to being predicted, the traffic data predicted value of Spatial Dimension is obtained, comprising:
According to the section at least two layers of road network of target road section upstream neighbor in current time previous predicted time interval
Historical traffic data, the traffic data to target road section in subsequent time predict that the traffic data for obtaining Spatial Dimension is pre-
Measured value.
5. urban highway traffic prediction technique according to claim 4, which is characterized in that described according to target road section upstream
Historical traffic data of the section in current time previous predicted time interval at least two layers of road network of neighbour, to target road
Section is predicted in the traffic data of subsequent time, obtains the traffic data predicted value of Spatial Dimension, comprising:
Calculate the upstream section of target road section and the Euclidean distance of target road section；The upstream section is target road section upstream neighbor
At least two layers of road network in section；
According to actual flow of the upstream section in the actual flow at current time and current time previous predicted time interval, with
The inverse of the Euclidean distance of upstream section and target road section calculates the volume forecasting of the subsequent time of target road section as weight
Value；
According to actual speed of the upstream section in the actual speed at current time and current time previous predicted time interval, with
The inverse of the Euclidean distance of upstream section and target road section calculates the prediction of speed of the subsequent time of target road section as weight
Value.
6. urban highway traffic prediction technique according to claim 1, which is characterized in that described to calculate the time dimension
Traffic data predicted value and the Spatial Dimension traffic data predicted value weighted average, obtain target road section next
The traffic data predicted value at moment, comprising:
Calculate the flow related coefficient of time dimension and the flow related coefficient of Spatial Dimension；
Using the flow related coefficient of the flow related coefficient of calculated time dimension and Spatial Dimension as weight, the time is calculated
The weighted average of the traffic prediction value of the traffic prediction value and Spatial Dimension of dimension obtains target road section in the stream of subsequent time
Measure predicted value；
Calculate the velocity correlation coefficint of time dimension and the velocity correlation coefficint of Spatial Dimension；
Using the velocity correlation coefficint of the velocity correlation coefficint of calculated time dimension and Spatial Dimension as weight, the time is calculated
The weighted average of the rate predictions of the rate predictions and Spatial Dimension of dimension obtains target road section in the speed of subsequent time
Spend predicted value.
7. urban highway traffic prediction technique according to claim 6, which is characterized in that the stream for calculating time dimension
Measure the flow related coefficient of related coefficient and Spatial Dimension, comprising:
According to the traffic prediction value and flow actual value of the time dimension in top n period at current time, pass through Pearson came phase relation
Number, calculates the flow related coefficient of time dimension and the flow related coefficient of Spatial Dimension, wherein N is the natural number greater than 0；
The velocity correlation coefficint of the velocity correlation coefficint for calculating time dimension and Spatial Dimension, comprising: according to current time
The rate predictions and speed actual value of the time dimension in top n period calculate time dimension by Pearson correlation coefficient
The velocity correlation coefficint of velocity correlation coefficint and Spatial Dimension, wherein N is the natural number greater than 0.
8. a kind of urban highway traffic prediction meanss characterized by comprising
Traffic data obtains module, for obtaining urban road history of the target road section before current time in predetermined time length
Traffic data；The historical traffic data includes data on flows and speed data；
Time dimension prediction module, for according at least two different chronon dimensions and the historical traffic data, to mesh
Mark section is predicted in the traffic data of subsequent time, obtains the traffic data predicted value of time dimension；
Spatial Dimension prediction module, for the upstream section according to target road section in current time previous predicted time interval
Historical traffic data, the traffic data to target road section in subsequent time predict that the traffic data for obtaining Spatial Dimension is pre-
Measured value；
Temporal-spatial fusion module, for calculating the traffic data predicted value of the time dimension and the traffic data of the Spatial Dimension
The weighted average of predicted value obtains target road section in the traffic data predicted value of subsequent time.
9. a kind of electronic equipment, which is characterized in that the electronic equipment includes: processor and memory, wherein memory is used for
Store executable program code；Processor is run and executable journey by reading the executable program code stored in memory
The corresponding program of sequence code, for executing urban highway traffic prediction technique described in aforementioned any claim.
10. a kind of computer readable storage medium, which is characterized in that the computer-readable recording medium storage have one or
Multiple one or more of programs of program can be executed by one or more processor, to realize aforementioned any claim
The urban highway traffic prediction technique.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810497797.9A CN110517479B (en) | 2018-05-22 | 2018-05-22 | Urban road traffic prediction method and device and electronic equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810497797.9A CN110517479B (en) | 2018-05-22 | 2018-05-22 | Urban road traffic prediction method and device and electronic equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110517479A true CN110517479A (en) | 2019-11-29 |
| CN110517479B CN110517479B (en) | 2020-11-03 |
Family
ID=68622312
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810497797.9A Active CN110517479B (en) | 2018-05-22 | 2018-05-22 | Urban road traffic prediction method and device and electronic equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110517479B (en) |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111062746A (en) * | 2019-12-06 | 2020-04-24 | 微梦创科网络科技（中国）有限公司 | Advertisement flow estimation method and device and electronic equipment |
| CN111553737A (en) * | 2020-04-30 | 2020-08-18 | 北京金山安全软件有限公司 | Business time prediction method and device |
| CN111612243A (en) * | 2020-05-18 | 2020-09-01 | 湖南大学 | Traffic speed prediction method, system and storage medium |
| CN111667694A (en) * | 2020-06-22 | 2020-09-15 | 河北工业大学 | Short-term traffic flow prediction method based on improved DTW-KNN |
| CN111724601A (en) * | 2020-06-30 | 2020-09-29 | 西安电子科技大学 | Motor vehicle traffic flow prediction method suitable for data loss |
| CN112102610A (en) * | 2020-07-27 | 2020-12-18 | 清华大学深圳国际研究生院 | Traffic flow prediction method |
| CN112907969A (en) * | 2021-02-02 | 2021-06-04 | 中国科学院计算技术研究所 | Method and system for predicting road traffic flow |
| CN113592136A (en) * | 2021-06-18 | 2021-11-02 | 阿波罗智联(北京)科技有限公司 | Traffic data prediction method and device and electronic equipment |
| CN113689699A (en) * | 2021-08-25 | 2021-11-23 | 浙江数智交院科技股份有限公司 | Traffic flow prediction method and device, electronic equipment and storage medium |
| CN114677844A (en) * | 2022-04-29 | 2022-06-28 | 黄小兵 | Intelligent traffic resource allocation service system and method |
| CN115440038A (en) * | 2022-08-31 | 2022-12-06 | 青岛海信网络科技股份有限公司 | Traffic information determination method and electronic equipment |
Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101488284A (en) * | 2008-01-16 | 2009-07-22 | 闵万里 | Intelligent management system for road traffic condition instant prediction |
| 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 |
| EP2590151A1 (en) * | 2011-11-03 | 2013-05-08 | Deutsche Telekom AG | A framework for the systematic study of vehicular mobility and the analysis of city dynamics using public web cameras |
| CN103903422A (en) * | 2012-12-24 | 2014-07-02 | 北京掌城科技有限公司 | Multi-source traffic information fusion method |
| CN104064023A (en) * | 2014-06-18 | 2014-09-24 | 银江股份有限公司 | Dynamic traffic flow prediction method based on space-time correlation |
| CN105321356A (en) * | 2015-12-08 | 2016-02-10 | 上海电机学院 | 2-dimensional flow prediction based intelligent urban traffic-flow prediction method |
| CN106128100A (en) * | 2016-06-30 | 2016-11-16 | 华南理工大学 | A kind of short-term traffic flow forecast method based on Spark platform |
| CN104821082B (en) * | 2015-04-29 | 2017-01-18 | 电子科技大学 | Short-time traffic flow prediction method based on integrated evaluation |
| CN106652443A (en) * | 2016-10-21 | 2017-05-10 | 长春理工大学 | Longitudinal and transverse similar highway short-time traffic prediction method |
| CN106940929A (en) * | 2016-01-04 | 2017-07-11 | 杭州海康威视数字技术股份有限公司 | Traffic data Forecasting Methodology and device |
| JP2017220000A (en) * | 2016-06-07 | 2017-12-14 | 日本電信電話株式会社 | Flow rate prediction device, pattern estimation device, flow rate prediction method, pattern estimation method and program |
-
2018
- 2018-05-22 CN CN201810497797.9A patent/CN110517479B/en active Active
Patent Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101488284A (en) * | 2008-01-16 | 2009-07-22 | 闵万里 | Intelligent management system for road traffic condition instant prediction |
| 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 |
| EP2590151A1 (en) * | 2011-11-03 | 2013-05-08 | Deutsche Telekom AG | A framework for the systematic study of vehicular mobility and the analysis of city dynamics using public web cameras |
| CN103903422A (en) * | 2012-12-24 | 2014-07-02 | 北京掌城科技有限公司 | Multi-source traffic information fusion method |
| CN104064023A (en) * | 2014-06-18 | 2014-09-24 | 银江股份有限公司 | Dynamic traffic flow prediction method based on space-time correlation |
| CN104821082B (en) * | 2015-04-29 | 2017-01-18 | 电子科技大学 | Short-time traffic flow prediction method based on integrated evaluation |
| CN105321356A (en) * | 2015-12-08 | 2016-02-10 | 上海电机学院 | 2-dimensional flow prediction based intelligent urban traffic-flow prediction method |
| CN106940929A (en) * | 2016-01-04 | 2017-07-11 | 杭州海康威视数字技术股份有限公司 | Traffic data Forecasting Methodology and device |
| JP2017220000A (en) * | 2016-06-07 | 2017-12-14 | 日本電信電話株式会社 | Flow rate prediction device, pattern estimation device, flow rate prediction method, pattern estimation method and program |
| CN106128100A (en) * | 2016-06-30 | 2016-11-16 | 华南理工大学 | A kind of short-term traffic flow forecast method based on Spark platform |
| CN106652443A (en) * | 2016-10-21 | 2017-05-10 | 长春理工大学 | Longitudinal and transverse similar highway short-time traffic prediction method |
Non-Patent Citations (1)
| Title |
|---|
| 许伦辉等: "基于特性和影响因素分析的短时交通流预测", 《广西师范大学学报：自然科学版》 * |
Cited By (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111062746A (en) * | 2019-12-06 | 2020-04-24 | 微梦创科网络科技（中国）有限公司 | Advertisement flow estimation method and device and electronic equipment |
| CN111553737B (en) * | 2020-04-30 | 2023-08-22 | 北京金山安全软件有限公司 | Business time prediction method and device |
| CN111553737A (en) * | 2020-04-30 | 2020-08-18 | 北京金山安全软件有限公司 | Business time prediction method and device |
| CN111612243A (en) * | 2020-05-18 | 2020-09-01 | 湖南大学 | Traffic speed prediction method, system and storage medium |
| CN111612243B (en) * | 2020-05-18 | 2023-11-10 | 湖南大学 | Traffic speed prediction method, system and storage medium |
| CN111667694A (en) * | 2020-06-22 | 2020-09-15 | 河北工业大学 | Short-term traffic flow prediction method based on improved DTW-KNN |
| CN111724601A (en) * | 2020-06-30 | 2020-09-29 | 西安电子科技大学 | Motor vehicle traffic flow prediction method suitable for data loss |
| CN112102610A (en) * | 2020-07-27 | 2020-12-18 | 清华大学深圳国际研究生院 | Traffic flow prediction method |
| CN112907969A (en) * | 2021-02-02 | 2021-06-04 | 中国科学院计算技术研究所 | Method and system for predicting road traffic flow |
| CN112907969B (en) * | 2021-02-02 | 2022-04-22 | 中国科学院计算技术研究所 | Method and system for predicting road traffic flow |
| CN113592136A (en) * | 2021-06-18 | 2021-11-02 | 阿波罗智联(北京)科技有限公司 | Traffic data prediction method and device and electronic equipment |
| CN113689699A (en) * | 2021-08-25 | 2021-11-23 | 浙江数智交院科技股份有限公司 | Traffic flow prediction method and device, electronic equipment and storage medium |
| CN114677844B (en) * | 2022-04-29 | 2022-11-22 | 立昂技术股份有限公司 | Intelligent traffic resource allocation service system and method |
| CN114677844A (en) * | 2022-04-29 | 2022-06-28 | 黄小兵 | Intelligent traffic resource allocation service system and method |
| CN115440038A (en) * | 2022-08-31 | 2022-12-06 | 青岛海信网络科技股份有限公司 | Traffic information determination method and electronic equipment |
| CN115440038B (en) * | 2022-08-31 | 2023-11-03 | 青岛海信网络科技股份有限公司 | Traffic information determining method and electronic equipment |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110517479B (en) | 2020-11-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110517479A (en) | A kind of urban highway traffic prediction technique, device and electronic equipment | |
| CN109376969A (en) | City fining population distribution dynamic prediction method and device based on deep learning | |
| CN109285346A (en) | A kind of city road net traffic state prediction technique based on key road segment | |
| CN109146204A (en) | A kind of wind power plant booster stations automatic addressing method of comprehensiveestimation | |
| CN110414732B (en) | Travel future trajectory prediction method and device, storage medium and electronic equipment | |
| CN105843942A (en) | Urban flood prevention decision support system based on big data technique | |
| CN105551239B (en) | travel time prediction method and device | |
| CN109376935A (en) | A kind of bus passenger flow neural network based combination forecasting method at times | |
| CN112200351A (en) | Urban area passenger flow volume prediction method based on mobile phone signaling data | |
| Mazloumi et al. | Exploring the value of traffic flow data in bus travel time prediction | |
| CN109448361A (en) | Resident's traffic trip volume forecasting system and its prediction technique | |
| CN107193060B (en) | A kind of multipath Typhoon Storm Surge Over method for quick predicting and system | |
| CN109544690A (en) | Shared bicycle trip influence factor recognition methods, system and storage medium | |
| CN110110243A (en) | A kind of historical track destination prediction technique based on echo state network | |
| Ahmed | Modelling spatio-temporal urban land cover growth dynamics using remote sensing and GIS techniques: A case study of Khulna City | |
| CN109360421A (en) | A kind of traffic information predicting method and device, electric terminal based on machine learning | |
| CN110335067A (en) | A kind of transaction prediction model construction method and device | |
| CN108734958A (en) | A kind of traffic speed prediction technique | |
| CN109635859A (en) | It is resident number combination forecasting method in Passenger Transport Hub region based on numerical value and trend k neighbour | |
| KR20090032461A (en) | Method of estimating salt contamination | |
| Murat et al. | Modeling vehicle delays at signalized junctions: Artificial neural networks approach | |
| CN106781508B (en) | Short-time Traffic Flow Forecasting Methods based on multiple phase space under a kind of Spark environment | |
| CN109558961A (en) | Determine method and system, storage medium, processor and the device of location information | |
| KR102359902B1 (en) | Crossroads LOS Prediction Method Based on Big Data and AI, and Storage Medium Having the Same | |
| CN111008730B (en) | Crowd concentration prediction model construction method and device based on urban space structure |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |