CN104464291B - Traffic flow predicting method and system - Google Patents
Traffic flow predicting method and system Download PDFInfo
- Publication number
- CN104464291B CN104464291B CN201410742725.8A CN201410742725A CN104464291B CN 104464291 B CN104464291 B CN 104464291B CN 201410742725 A CN201410742725 A CN 201410742725A CN 104464291 B CN104464291 B CN 104464291B
- Authority
- CN
- China
- Prior art keywords
- data
- traffic flow
- section
- historical
- forecast model
- 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.)
- Active
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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
Abstract
The embodiment of the invention provides a traffic predicting method and system. The method comprises the steps that current traffic flow data are acquired, wherein the current traffic flow data comprise current traffic flow data of a road segment to be measured and current traffic flow data road segments related to the road segment to be measured; a first predicting model is determined; the current traffic flow data are input to the first predicting model, and a predicting traffic flow value of the road segment to be measured is obtained, wherein the first predicting model carries out discrete processing on historical traffic flow data by acquiring the historical traffic flow data to obtain discrete historical data, a space-time association rule is obtained by mining the discrete historical data, a set of the road segments related to the road segment to be measured is obtained through the space-time association rule, and the historical traffic flow data of the set of the road segments are generated in a training mode through a preset method.
Description
Technical field
The present invention relates to traffic flow forecasting field, more particularly to a kind of traffic flow forecasting method and system.
Background technology
With the continuous development of social economy and transportation, the traffic problems such as traffic congestion increasingly show especially out,
Become the problem of global common concern.Prediction for traffic flow is not only the basis of Urban traffic control and route guidance, or solution
Certainly one of key technology of congestion in road problem.If the flow motor on each branch road in the network of communication lines can accurately be predicted, that
We reasonably can be optimized to traffic flow with planing method, so that the utilization rate of road reaches maximum,
Part congestion problems can be solved.
At present, carry out predicting traffic flow amount frequently with prediction of short-term traffic volume model, predict mould compared with the short-term traffic flow of getting up early
Type has: autoregression model (ar), moving average model (ma), autoregressive moving-average model (arma), history averaging model
(ha) and box-cox model etc., with the development in this field, Forecasting Methodology constantly tends to accurate, occur in that many more complicated,
The higher forecast model of precision, generally can be divided into two classes: a class is with traditional mathematics sides such as mathematical statisticss and calculus
Forecast model based on method, specifically includes that time series models, Kalman filter model, Partial Linear Models etc.；Equations of The Second Kind
Be with modern science and technology and method (as analogue technique, neutral net, analogue technique) be main research meanses formed short
Phase forecast model, this kind of method does not pursue proper mathematical derivation and clear and definite physical significance, more payes attention to and reality
The closeness of fit of traffic flow, this kind of method mainly includes nonparametric Regression Model, arima algorithm, is based on wavelet theory
Method, analysis of spectrum and multiple hybrid model for short-term load forecasting related to neutral net etc..
Carry out predicting traffic flow amount by using short-term traffic flow forecast model, can effectively predict section to be measured and exist
The concrete numerical value of traffic flow in certain future time, to section to be measured in certain future time traffic flow carry out accurately pre-
Survey, but, due to features such as the non-linear, complexity of traffic system and uncertainties, existing prediction of short-term traffic volume model holds
Interference factor is affected immediately to be easily subject to the external world, because the impact of external world's interference factor immediately, using existing prediction of short-term traffic volume
Model carry out predicting traffic flow amount so as to get predicting traffic flow value Stability and veracity low.
Content of the invention
In view of this, the embodiment of the present invention provides a kind of traffic flow forecasting method and system, to solve in prior art
Because the impact of external world's interference factor immediately, carry out predicting traffic flow amount using existing prediction of short-term traffic volume model so as to get
The low problem of the Stability and veracity of predicting traffic flow value.
For achieving the above object, the following technical scheme of embodiment of the present invention offer:
A kind of traffic flow forecasting method, comprising:
Obtain current traffic flow data, described current traffic flow data includes the current traffic flow number in section to be measured
According to and with section to be measured have relatedness section Current traffic flow data；
Determine the first forecast model；
By the first forecast model described in described current traffic flow data input, obtain the predicting traffic flow amount in section to be measured
Value；
Wherein, described first forecast model passes through to obtain historical traffic flows data, historical traffic stream described in discrete processes
Amount data, obtains discretization historical data, described discretization historical data is carried out with data mining and obtains space time correlation rule, by
Described space time correlation rule obtains the section set having relatedness with section to be measured, by the historical traffic fluxion of described section set
Generate according to being trained by preordering method.
Wherein, described determination the first forecast model includes:
Current traffic flow data described in discrete processes, obtains discretization current data；
Described discretization current data is inputted the second forecast model, obtains traffic flow trend；
According to corresponding between described section to be measured identification information, current time information and traffic flow trend and the first forecast model
The first forecast model that relation determination is matched with the second forecast model.
Wherein, described second forecast model passes through to obtain historical traffic flows data, historical traffic described in sliding-model control
Data on flows, historical traffic flow data after sliding-model control is carried out the space time correlation rule composition that data mining obtains.
Wherein, according to current traffic flow data described in following steps discrete processes:
The current traffic flow data normalization obtaining is interval interior to [0,1]；
Discretization is carried out to current traffic flow data after described normalized, obtains discretization current data；
Wherein, according to historical traffic flows data described in following steps discrete processes:
The historical traffic circuit-switched data of acquisition is normalized to [0,1] interval interior；
Discretization is carried out to historical traffic flows data after described normalized, obtains discretization historical data.
Wherein, data mining is carried out by evolutionary programming algorithm to described discretization historical data and obtain space time correlation rule
Then；
Wherein, data mining is carried out by evolutionary programming algorithm to described discretization historical data and obtain space time correlation rule
Including:
Determine the number m in section in discretization historical data, generate the individuality with m gene；
Take the individual composition initial population not less than predetermined number, and described initial population is calculated according to discretization historical data
Each individual fitness value in body；
Swarm Evolution is carried out to described initial population according to each individual fitness value in described initial population；
Qualified correlation rules all in each individuality after Swarm Evolution are exported, is obtained space time correlation rule.
Wherein, described generation has the individual of m gene and includes:
Generate the determination gene that a span is [- t, -1], t is traffic flow data in historical traffic flows data
The quantity of sampling time interval；
Generate the random gene that m-1 span is [0, t], obtain the individuality with m gene.
Wherein, described colony is carried out to described initial population according to each individual fitness value in described initial population enter
Change and include:
Randomly choose two individualities in described initial population, select the larger individual entrance of wherein fitness value of future generation, directly
To selecting, predetermined number is individual to form new colony；
Randomly choose out predetermined number individuality and complete chromosomal chiasma operation from described new colony；
Randomly choose out predetermined number individuality and complete chromosome disorder operation from described new colony；
Repeat new colony and generate operation, chromosomal chiasma operation and chromosome disorder operation, until colony's fitness
Value convergence.
Wherein, described current traffic flow data and described historical traffic flows data are and have filtered out abnormal number
According to, and the abnormal data filtering out was carried out with the data revised.
Wherein, described and section to be measured tool relatedness section Current traffic stream packets include:
Determine section to be measured and space time correlation rule；
Obtain the section having relatedness with section to be measured according to described section to be measured and space time correlation rule.
A kind of traffic flow forecasting system, comprising: acquisition module, model determining module, prediction module and forecast model life
Become module；Wherein,
Described acquisition module, for obtaining current traffic flow data, described current traffic flow data includes road to be measured
The current traffic flow data of section and the Current traffic flow data having relatedness section with section to be measured；
Described model determining module, for determining the first forecast model；
Described prediction module, for by the first forecast model described in described current traffic flow data input, obtaining to be measured
The predicting traffic flow value in section；
Described forecast model generation module, for obtaining historical traffic flows data, historical traffic described in sliding-model control
Data on flows, historical traffic flow data after sliding-model control is carried out data mining and obtains space time correlation rule, by described space-time
Correlation rule obtains the section set having relatedness with section to be measured, the historical traffic flow data of described section set is passed through pre-
Determine method training and generate the first forecast model.
Wherein, described model determining module includes: forecast model determining unit and forecast model signal generating unit；Wherein,
Described forecast model determining unit, for current traffic flow data described in sliding-model control, obtains discretization and works as
Front data；Described discretization current data is inputted the second forecast model, obtains traffic flow trend；Become according to described traffic flow
The first forecast model that between gesture and the first forecast model, corresponding relation determination is matched with the second forecast model；
Described forecast model signal generating unit, for obtaining historical traffic flows data, historical traffic described in sliding-model control
Data on flows, historical traffic flow data after sliding-model control is carried out the space time correlation rule that data mining obtains, forms second
Forecast model.
Based on technique scheme, a kind of traffic flow forecasting method provided in an embodiment of the present invention and system, obtaining
After current traffic flow data, by determining the first forecast model, by the first of the current flows data input obtaining determination
Forecast model, then obtains the predicting traffic flow value in section to be measured from the first forecast model determining, wherein, described current
Traffic flow data includes the current traffic flow data in section to be measured and the Current traffic having relatedness section with section to be measured
Flow data, wherein, described first forecast model passes through to obtain historical traffic flows data, historical traffic flows described in discrete processes
Data, obtains discretization historical data, described discretization historical data is carried out with data mining and obtains space time correlation rule, by institute
State the section set that space time correlation rule obtains having relatedness with section to be measured, by the historical traffic flow data of described section set
Trained by preordering method and generate.Because in the traffic flow forecasting method that this transmission embodiment provides, the first prediction of use
Model is that history communication flow data obtains space time correlation rule through data mining, needed for being obtained by this moment correlation rule
The history communication flow data message in section obtained by preordering method training, data carried out to historical traffic data dig
Pick, by carrying out global optimum's search to mass data, obtains frequent item set, improves the accuracy of predicting traffic flow value.
Brief description
In order to be illustrated more clearly that the embodiment of the present invention or technical scheme of the prior art, below will be to embodiment or existing
Have technology description in required use accompanying drawing be briefly described it should be apparent that, drawings in the following description be only this
Inventive embodiment, for those of ordinary skill in the art, on the premise of not paying creative work, can also basis
The accompanying drawing providing obtains other accompanying drawings.
Fig. 1 is the flow chart of circulation method for predicting provided in an embodiment of the present invention；
Fig. 2 is the method flow determining the first forecast model in traffic flow forecasting method provided in an embodiment of the present invention
Figure；
Fig. 3 is the side of discrete processes current traffic flow data in traffic flow forecasting method provided in an embodiment of the present invention
Method flow chart；
Fig. 4 is the side of discrete processes historical traffic flows data in traffic flow forecasting method provided in an embodiment of the present invention
Method flow chart；
Fig. 5 is to discretization history in traffic flow forecasting method provided in an embodiment of the present invention by evolutionary programming algorithm
Data carries out the method flow diagram that data mining obtains space time correlation rule；
Fig. 6 is the method generating the individuality with m gene in traffic flow forecasting method provided in an embodiment of the present invention
Flow chart；
Fig. 7 is the adaptation in traffic flow forecasting method provided in an embodiment of the present invention according to each individuality in initial population
Angle value carries out the method flow diagram of Swarm Evolution to described initial population；
Fig. 8 is to obtain having relatedness section with section to be measured in traffic flow forecasting method provided in an embodiment of the present invention
Method flow diagram；
Fig. 9 is the system block diagram of traffic flow forecasting system provided in an embodiment of the present invention；
Figure 10 is the structural representation of model determining module 200 in traffic flow forecasting system provided in an embodiment of the present invention
Figure.
Specific embodiment
Below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is carried out clear, complete
Site preparation description is it is clear that described embodiment is only a part of embodiment of the present invention, rather than whole embodiments.It is based on
Embodiment in the present invention, it is every other that those of ordinary skill in the art are obtained under the premise of not making creative work
Embodiment, broadly falls into the scope of protection of the invention.
Fig. 1 is the flow chart of circulation method for predicting provided in an embodiment of the present invention, by entering to historical traffic data
Row data mining, the historical traffic flow data obtaining having relatedness section with section to be measured trains the first forecast model obtaining
Predicting traffic flow amount, improves the accuracy of predicting traffic flow value, and with reference to Fig. 1, this circulation method for predicting may include that
Step s100: obtain current traffic flow data, described current traffic flow data includes the current of section to be measured
Traffic flow data and the Current traffic flow data having relatedness section with section to be measured；
Wherein it is desired to explanation, when not having the section of relatedness with section to be measured then it is assumed that obtaining and road to be measured
The Current traffic flow data in section tool relatedness section be empty data it is also possible to think, when not and section to be measured has relatedness
During section, only include the current traffic flow data in section to be measured in the current traffic flow data of acquisition, and do not include and treat
This survey section has the Current traffic flow data in relatedness section.
By obtaining current communication flow data, current time letter can be determined from the current traffic flow data obtaining
Breath, section to be measured identification information and current traffic flow value data information, optionally, section to be measured identification information can be for treating
Survey title or the number information in section.
Optionally, can be thought in the title confirming section to be measured or numbering with the section treating this survey section tool relatedness
After information and the prior space time correlation rule to learn, being had with section to be measured of being obtained according to survey section and space time correlation rule is closed
The section of connection property.
Step s110: determine the first forecast model, described first forecast model passes through to obtain historical traffic flows data, from
Dissipate and process described historical traffic flows data, obtain discretization historical data, data is carried out to described discretization historical data and digs
Pick obtains space time correlation rule, is obtained the section set having relatedness with section to be measured by described space time correlation rule, will be described
The historical traffic flow data of section set is trained by preordering method and is generated；
Historical traffic flows data includes system unit time, current time, time interval quantity, section mark and hands over
The information such as through-current capacity, wherein, the system unit time is the time interval of data statisticss, can be by User Defined, generally with 5 points
Clock or 10 minutes are unit, and minima is 1 minute, and the traffic flow statistics less than a minute does not possess statistical significance.
For example, as shown in table 1, the historical traffic flows data of some day:
The historical traffic flows data form of table 1 some day
Wherein, the △ t in form is the system unit time, and t is time interval quantity, and optionally, t can be with value as 24*
60/ △ t, t are current time；blk_idiIdentify for section, wherein i is expressed as i+1 section, for example, the 1st section
Section is designated blk_id0, the section mark as blk_id in the 2nd section1；tfiFor traffic flow, i therein also is indicated as
I-th section.
Wherein it is desired to explanation, table 1 is only the historical traffic flows data form of some day, and the history obtaining is handed over
The data in the historical traffic flows data form of many days is included in through-current capacity data, optionally, the historical traffic stream of acquisition
Amount data in for 30 historical traffic flows data it may also be said to, obtain 30 historical traffic flows data forms.
Optionally, can be by certain moment t, certain section blk_idiTraffic flow be designated as tfi,t, tfi,tObtain shown in being
Historical traffic flows data, with reference to table 1 it can be seen that during 1 moment, i.e. t1During=△ t, the 1st section, i.e. blk_idiFor
blk_id0When traffic flow be 123, i.e. tf in table 10,1=123, you can to get historical traffic flows in Table 1
Data tf0,1=123, section to be measured identification information, historical time information can be determined from the historical traffic flows data obtaining
With historical traffic flows value data information.
Optionally, can be from flow collection system or the road obtaining designated area from the data on flows storehouse of business platform
Road historical traffic flows data.
Optionally, discrete processes historical traffic flows data can be, the historical traffic circuit-switched data normalization that first will obtain
Interval interior to [0,1], then discretization is carried out to historical traffic flows data after normalized, obtain discretization historical data.
Optionally, historical traffic flows data discrete can be tri- numerical value of l, m and h, represent traffic flow respectively and become
Gesture is low, medium and high.
Optionally, the historical traffic flows data of acquisition can be for having filtered out abnormal data and different to filter out
Regular data carried out the data revised.Using having filtered out abnormal data, and correction was carried out to the abnormal data filtering out
Historical traffic flows data, can avoid follow-up result is produced with the situation of noise；Speed, traffic flow, occupation rate number
According to the situation that zero or negative value occur；Traffic flow exceedes more than design discharge certain limit, or speed is more than detection range
Situation；The judgement relation of speed, occupation rate and vehicle commander's scope, that is, by speed, occupation rate, the product in traffic flow collection cycle, can
The Vehicle length extrapolated, off-limits situation for vehicle commander；Missing data etc. the occurrence of, further increase traffic
The Stability and veracity of volume forecasting.
Optionally, because traffic flow data has the characteristics that will not undergo mutation within the system unit time, therefore, can
The abnormal data filtering out to be modified using arest neighbors interpolation method.
Table 2 gives through data correction and the historical traffic flows data form after discrete processes:
The table 2 historical traffic flows data form through data correction and after discrete processes
Table 1 is mutually compareed with table 2 it is found that the data of Lycoperdon polymorphum Vitt shading is the historical traffic stream after revising in table 2
Amount data.
Optionally, data mining can be carried out by evolutionary programming algorithm to described discretization historical data and obtain space-time pass
Connection rule.Generally, data mining algorithm using string every in tables of data as an independent attribute, data line conduct
One data tuple, completes to excavate the correlation rule between two community sets by statistical analysiss.For traffic flow forecasting
Association rule mining and traditional data mining have two main differences: one, traffic data has spatiotemporal, and that is, correlation rule is not
Exist only between two property sets of synchronization, also can be present in not between two property sets in the same time；This space-time
Characteristic can lead to the growth of data-mining search space exponentially rank, can cause greatly to choose for prognoses system computing capability
War；2nd, the follow-up item of the correlation rule of forecasting traffic flow is fixing, blk_id such as to be predicted20In t=tjThe traffic flow in moment
Measure, the follow-up item of its correlation rule isWith
Therefore, complete the search of correlation rule using evolutionary programming algorithm, due to evolutionary programming algorithm high concurrent and
Probabilistic search feature, can effectively reduce the complexity of correlation rule；Meanwhile, because the space-time characterisation of traffic parameter can become quantity
The complexity of the increase data mining of level, can effectively improve data mining using the high concurrent based on evolutionary programming algorithm
Efficiency；Integrating parallel computing environment can reach real-time update forecast model.
Optionally, data mining is carried out by evolutionary programming algorithm to described discretization historical data and obtain space time correlation rule
Can be then first to determine the number m in section in discretization historical data, generate the individuality with m gene, wherein, in individuality
Each gene and section correspond, then take the individual composition initial population not less than predetermined number, and according to discretization history
Data calculates each individual fitness value in described initial population, the then adaptation according to each individuality in described initial population
Angle value carries out Swarm Evolution to described initial population, finally enters qualified correlation rules all in each individuality after Swarm Evolution
Row output, obtains space time correlation rule.
Optionally, determine the number m in section in discretization historical data, generate in the individuality with m gene, generation
M gene in individuality, is the determination gene of [- t, -1] including a span, the corresponding section of this gene, is to be measured
Section, wherein, t is the quantity of traffic flow data sampling time interval in historical traffic flows data；With m-1 span
For the random gene of [0, t], wherein, value be 0 the gene representation section corresponding with this gene unrelated with section to be measured, no
Participate in the calculating of space time correlation rule.
Optionally, can generate m-1 span using uniform random number generation method is the random of [0, t]
Gene.
For example, as shown in table 3, an individuality with 635 genes:
Table 3, an individuality with 635 genes
| 0 | 7 | … | -1 | 0 | … | 0 | 9 |
| Gene 1 | Gene 2 | Gene 78 | Gene 79 | Gene 634 | Gene 635 |
In table 3 it can be seen that this individuality has 635 genes altogether, individuality altogether has 635 gene representations and obtains
Discretization historical data in there are 635 sections；- 1 represents that the 78th corresponding section of gene is section to be measured, and pre-
Survey is the traffic flow data in the 1st moment；Other numerals, such as 7, represent between the 2nd corresponding section of gene and section to be measured
There is space time correlation relatedness, and associate the moment for the 7th moment.
According to table 3, correlation rule forerunner's attribute set of this individuality representative can be obtained, that is, section subset is { blk_
Id2, blk_id635 }, following 12 correlation rules that can be wherein contained:
R1:
R2:
……
R7:
R8:
……
R12:
Optionally, η nt individual composition initial population can be selected, represent the traffic flow shape in each moment for each section
State, as the follow-up item of space time correlation rule, wherein, η is population size control parameter, and optionally, the numerical value of η can be taken at
Between 10-100.
In genetic algorithm, to evaluate the good and bad degree of each individuality with the size of individual adaptation degree, thus determining its something lost
The size of biography chance, the generally negated negative value of fitness function, and with the fitness value of maximization colony for optimizing mesh
Mark.
Optionally, can adopt and each correlation rule is weighed based on support support and confidence level confidence
Quality.Wherein, emphasis characterizes the separating capacity of rule using confidence level, takes into account the ability of aggregation of correlation rule simultaneously.False
If whole historical traffic flow data table a total of k bar record, by all data recorders in convenience data storehouse taking r1 as a example
Calculation support and confidence level:
When r1 meets sup (r1) > θ sup and conf (r1) > θ conf simultaneously, it is one and qualified pass that r1 just will be considered that
Connection rule.Optionally, the span of θ sup can be 0.25-0.6；The span of θ conf can be 0.6-0.9.
Optionally, an individual all correlation rules support and confidence level can be calculated using a fitness function
Weighted sum, as follows:
Wherein, ω 1+ ω 2=1,0 < ω 1 < 1,0 < ω 2 < 1, ω 2 < ω 1；G represents to be currently g for population.
Optionally, fitness numerical value in this two individualities can be selected by randomly choosing two individualities in initial population
Larger individual entrance is of future generation, until selecting, predetermined number is individual to form new colony, then from the new colony obtaining at random
Select predetermined number individuality and complete chromosomal chiasma operation, randomly choose out predetermined number individuality from the new colony obtaining complete
Become chromosome disorder operation, repeat new colony and generate operation, chromosomal chiasma operation and chromosome disorder operation, until group
Body fitness value is restrained, and to carry out the Swarm Evolution of initial population.
Optionally, the number of the new individual in population selected can be η nt.
Optionally, the individual amount that can randomly choose out total individual number mesh 40%-70% from the new colony obtaining is complete
Become chromosomal chiasma operation.
Optionally, the individual amount that can randomly choose out total individual number mesh 0.1%-10% from the new colony obtaining is complete
Become chromosome disorder operation.
Optionally, colony's fitness value convergence formula can be:
| ∑ fitness (g, indj)-∑ fitness (g-1, indj) | ＜ ε；
After the space time correlation rule being obtained by data mining, found out using space time correlation rule and have with front section to be measured
The section set of relatedness, i.e. the preposition community set of all correlation rule subsets, by the historical traffic flows of this section set
Data sets up corresponding first forecast model as training data.
Optionally, because traffic flow data belongs to nonstationary time series, therefore, the first forecast model of training is permissible
For arima (p, d, q) model.
Optionally, with ri | tf78,1=l } as a example, training obtain obtaining during arima (p, d, q) model parameter p to be estimated,
The method of d and q may include that
Step a): obtain historical traffic flows data tf78,1, tf78,0..., tf78,1-p；tf2,7, tf2,6..., tf2,7-p,
And tf635,9, tf635,8..., tf635,9-p；
Step b): the traffic flow data of acquisition is converted to traffic flow differential data；
Step c): determine parameter d using single method of inspection；
Step d): parameter p is determined according to training data and parameter d；
Step e): by ar (∞), parameter d and p determine parameter q.
Optionally, with ri | tf78,1=l } as a example, the traffic flow data of acquisition is converted to traffic flow differential data
May include that
δtfI, t=tfI, t-tfI, t-1；
δ2tfI, t=δ tfI, t-δtfI, t-1；
……
Wherein, δ is first difference operator；δ2Position 2 order difference operators；Check the mark operator for d rank with δ d；α i is certainly
Regression parameter；δ j is rolling average parameter；P, d and q are model parameter to be estimated；ε is white noise.
Different sections, when different, all correspond to the first different predictions according to different traffic flow trend
Model, therefore, is trained by preordering method by the historical traffic flow data of the section set being had relatedness with section to be measured and generates
The first forecast model be multiple first forecast models, need before using the first forecast model predicting traffic flow value, right
Need the first forecast model using to be selected, determine that the current traffic flow data that the determination in section to be measured obtains is made
First forecast model, just can obtain section to be measured after current flows data input first forecast model that will obtain
Predicting traffic flow value.
Step s120: described current flows data input first forecast model obtains the prediction traffic in section to be measured
Flow value.
Based on technique scheme, a kind of traffic flow forecasting method provided in an embodiment of the present invention and system, obtaining
After current traffic flow data, by determining the first forecast model, by the first of the current flows data input obtaining determination
Forecast model, then obtains the predicting traffic flow value in section to be measured from the first forecast model determining, wherein, described current
Traffic flow data includes the current traffic flow data in section to be measured and the Current traffic having relatedness section with section to be measured
Flow data, wherein, described first forecast model passes through to obtain historical traffic flows data, historical traffic flows described in discrete processes
Data, obtains discretization historical data, described discretization historical data is carried out with data mining and obtains space time correlation rule, by institute
State the section set that space time correlation rule obtains having relatedness with section to be measured, by the historical traffic flow data of described section set
Trained by preordering method and generate.Because in the traffic flow forecasting method that this transmission embodiment provides, the first prediction of use
Model is that history communication flow data obtains space time correlation rule through data mining, needed for being obtained by this moment correlation rule
The history communication flow data message in section obtained by preordering method training, data carried out to historical traffic data dig
Pick, by carrying out global optimum's search to mass data, obtains frequent item set, improves the accuracy of predicting traffic flow value.
Optionally, Fig. 2 shows in the traffic flow forecasting method that point inventive embodiments provide and determines the first forecast model
Method flow diagram, with reference to Fig. 2, this determines that the method for the first forecast model may include that
Step s200: current traffic flow data described in discrete processes, obtain discretization current data；
Optionally, discrete processes historical traffic flows data can be, the historical traffic circuit-switched data normalization that first will obtain
Interval interior to [0,1], then discretization is carried out to historical traffic flows data after normalized, obtain discretization historical data.
Step s210: described discretization current data is inputted the second forecast model, obtains traffic flow trend, described the
Two forecast models pass through to obtain historical traffic flows data, and historical traffic flows data described in sliding-model control, at discretization
After reason, historical traffic flow data carries out the space time correlation rule composition that data mining obtains；
Second forecast model carries out, by historical traffic flow data after sliding-model control, the space time correlation rule that data mining obtains
Then collection is combined into, and each section is owned by several independent space time correlation rules subset in moment t, if by historical traffic stream
Amount data discrete is tri- numerical value of l, m and h, then each section is owned by 3 independent space time correlation rules subset in moment t.
Taking in table 3 as a example, section blk_id78, its forecast model is by { ri}、{riAnd
{riThree correlation rule subset compositions, wherein,
…；
…；
…；
Traffic flow dataWith correlation rule subset ri | tf78,1=l } in certain correlation rule rir matching degree such as
Under:
Traffic flow data tftAs follows with the matching degree of correlation rule subset { ri }:
Wherein, | ri | represents correlation rule quantity in correlation rule subset.
Model prediction resultRealized by maximizing correlation rule matching degree:
Optionally, if the historical traffic flows data of acquisition and current traffic flow data are carried out discrete processes, will
Data discrete is tri- numerical value of l, m and h, then can obtain traffic flow trend by the second forecast model can be low, neutralization
High.
Step s220: determined and the second forecast model according to corresponding relation between described traffic flow trend and the first forecast model
The first forecast model matching.
Because different sections, when different, different first all be correspond to according to different traffic flow trend
Forecast model, when obtaining current traffic flow data, just can learn mark and the current time in section to be measured, therefore, logical
After crossing the second forecast model, after learning traffic flow trend, just can determine that the first required forecast model.
Optionally, Fig. 3 shows discrete processes Current traffic in traffic flow forecasting method provided in an embodiment of the present invention
The method flow diagram of data on flows, with reference to Fig. 3, the method for this discrete processes current traffic flow data may include that
Step s300: the current traffic flow data normalization obtaining is interval interior to [0,1]；
Optionally, historical traffic circuit-switched data tf that will obtainI, tBeing normalized to [0,1] interval formula can be:
Mintf=min { tfi}；
Maxtf=max { tfi}；
Wherein, tfI, t' for being normalized to the historical traffic flows data behind [0,1] interval.
Step s310: discretization is carried out to current traffic flow data after described normalized, obtains discretization current
Data.
Optionally, discretization is carried out to historical traffic flows data after normalized, obtain discretization historical data
Formula can be:
Wherein,Traffic flow data for discretization；μiFor being numbered blk_idiSection history average traffic
Flow；σiFor being numbered blk_idiSection historical traffic flows standard deviation；θlAnd θhIt is discretization threshold value；αiFor adjusting
Weighted value, its numerical value can be adjusted according to practical situation, optionally, can say αiNumerical value be set between 0.25-0.5.
Optionally, Fig. 4 shows discrete processes historical traffic in traffic flow forecasting method provided in an embodiment of the present invention
The method flow diagram of data on flows, with reference to Fig. 4, the method for this discrete processes historical traffic flows data may include that
Step s400: the historical traffic circuit-switched data of acquisition is normalized to [0,1] interval interior；
Step s410: discretization is carried out to historical traffic flows data after described normalized, obtains discretization history
Data.
Wherein, historical traffic flows data is carried out with discrete processes and discrete processes are carried out to current traffic flow data is
Processed using identical data formula.
Optionally, Fig. 5 shows and passes through evolutionary programming algorithm in traffic flow forecasting method provided in an embodiment of the present invention
Discretization historical data is carried out with the method flow diagram that data mining obtains space time correlation rule, with reference to Fig. 5, by genetic evolution
Algorithm discretization historical data is carried out data mining obtain space time correlation rule method may include that
Step s500: determine the number m in section in discretization historical data, generate the individuality with m gene；
Wherein, in m gene in the individuality of generation, it is the determination gene of [- t, -1] including a span, this base
Because corresponding section is section to be measured, and m-1 span is the random gene of [0, t], and wherein, t is historical traffic stream
The quantity of traffic flow data sampling time interval in amount data, value is the 0 gene representation section corresponding with this gene
Unrelated with section to be measured, it is not involved in the calculating of space time correlation rule.
Step s510: take the individual composition initial population not less than predetermined number, and calculated according to discretization historical data
Each individual fitness value in described initial population；
In genetic algorithm, to evaluate the good and bad degree of each individuality with the size of individual adaptation degree, thus determining its something lost
The size of biography chance, the generally negated negative value of fitness function, and with the fitness value of maximization colony for optimizing mesh
Mark.
Optionally, η nt individual composition initial population can be selected, represent the traffic flow shape in each moment for each section
State, as the follow-up item of space time correlation rule, wherein, η is population size control parameter, and optionally, the numerical value of η can be taken at
Between 10-100.
Step s520: colony is carried out to described initial population according to each individual fitness value in described initial population and enters
Change；
Optionally, fitness numerical value in this two individualities can be selected by randomly choosing two individualities in initial population
Larger individual entrance is of future generation, until selecting, predetermined number is individual to form new colony, then from the new colony obtaining at random
Select predetermined number individuality and complete chromosomal chiasma operation, randomly choose out predetermined number individuality from the new colony obtaining complete
Become chromosome disorder operation, repeat new colony and generate operation, chromosomal chiasma operation and chromosome disorder operation, until group
Body fitness value is restrained, and to carry out the Swarm Evolution of initial population.
Step s530: qualified correlation rules all in each individuality after Swarm Evolution are exported, obtains space-time and close
Connection rule.
Optionally, can adopt and each correlation rule is weighed based on support support and confidence level confidence
Quality.Wherein, emphasis characterizes the separating capacity of rule using confidence level, takes into account the ability of aggregation of correlation rule simultaneously.
Taking the r1 obtaining in table 3 as a example, only when meeting sup (r1) > θ sup and conf (r1) > θ conf simultaneously,
Will be considered that it is that r1 is and qualified correlation rule.
Optionally, the span of θ sup can be 0.25-0.6；The span of θ conf can be 0.6-0.9.
Optionally, Fig. 6 shows to generate have m gene in traffic flow forecasting method provided in an embodiment of the present invention
Individual method flow diagram, with reference to Fig. 6, the method that this generation has the individuality of m gene may include that
Step s600: generate the determination gene that a span is [- t, -1]；
Wherein, t is the quantity of traffic flow data sampling time interval in historical traffic flows data.
The span generating is section to be measured for the corresponding section of the determination gene of [- t, -1].
Step s610: generate the random gene that m-1 span is [0, t].
Optionally, can generate m-1 span using uniform random number generation method is the random of [0, t]
Gene.
Optionally, Fig. 7 shows in traffic flow forecasting method provided in an embodiment of the present invention according to every in initial population
Individual fitness value carries out the method flow diagram of Swarm Evolution to described initial population, and with reference to Fig. 7, this is according to initial population
In each individual fitness value Swarm Evolution is carried out to described initial population method may include that
Step s700: randomly choose two individualities in described initial population, select the larger individual entrance of wherein fitness value
The next generation, until selecting, predetermined number is individual to form new colony；
Optionally, the number of the new individual in population selected can be η nt.
Step s710: randomly choose out predetermined number individuality from described new colony and complete chromosomal chiasma operation；
Optionally, the individual amount that can randomly choose out total individual number mesh 40%-70% from the new colony obtaining is complete
Become chromosomal chiasma operation.
Step s720: randomly choose out predetermined number individuality from described new colony and complete chromosome disorder operation；
Optionally, the individual amount that can randomly choose out total individual number mesh 0.1%-10% from the new colony obtaining is complete
Become chromosome disorder operation.
Step s730: repeat new colony and generate operation, chromosomal chiasma operation and chromosome disorder operation, until group
Body fitness value is restrained.
Optionally, colony's fitness value convergence formula can be:
| ∑ fitness (g, indj)-∑ fitness (g-1, indj) | ＜ ε
Optionally, Fig. 8 shows and obtains in traffic flow forecasting method provided in an embodiment of the present invention and section to be measured tool
The method flow diagram in relatedness section, with reference to Fig. 8, the method that this obtains having relatedness section with section to be measured may include that
Step s800: determine section to be measured and space time correlation rule；
Step s810: obtain the section having relatedness with section to be measured according to described section to be measured and space time correlation rule.
Traffic flow forecasting method provided in an embodiment of the present invention, the first forecast model of use is history communication flow number
Obtain space time correlation rule, the history communication flow in the section needed for being obtained by this moment correlation rule according to through data mining
Data message is obtained by preordering method training, carries out data mining to historical traffic data, by carrying out to mass data
Global optimum is searched for, and obtains frequent item set, improves the accuracy of predicting traffic flow value.
Below traffic flow forecasting system provided in an embodiment of the present invention is introduced, traffic flow described below is pre-
Examining system and above-described traffic flow forecasting method can be mutually to should refer to.
Fig. 9 is the system block diagram of traffic flow forecasting system provided in an embodiment of the present invention, with reference to Fig. 9, this traffic flow
Prognoses system may include that acquisition module 100, model determining module 200, prediction module 300 and forecast model generation module
400；Wherein,
Acquisition module 100, for obtaining current traffic flow data, described current traffic flow data includes section to be measured
Current traffic flow data and with section to be measured have relatedness section Current traffic flow data；
Model determining module 200, for determining the first forecast model；
Prediction module 300, for by the first forecast model described in described current traffic flow data input, obtaining road to be measured
The predicting traffic flow value of section；
Forecast model generation module 400, for obtaining historical traffic flows data, historical traffic stream described in sliding-model control
Amount data, historical traffic flow data after sliding-model control is carried out data mining and obtains space time correlation rule, closed by described space-time
Connection rule obtains the section set having relatedness with section to be measured, the historical traffic flow data of described section set is passed through predetermined
Method training generates the first forecast model.
Optionally, Figure 10 shows model determining module 200 in traffic flow forecasting system provided in an embodiment of the present invention
Structural representation, with reference to Figure 10, this model determining module 200 may include that forecast model determining unit 210 and forecast model
Signal generating unit 220；Wherein,
Forecast model determining unit 210, for current traffic flow data described in sliding-model control, obtains discretization current
Data；Described discretization current data is inputted the second forecast model, obtains traffic flow trend；According to described traffic flow trend
The first forecast model that corresponding relation determination is matched with the second forecast model and between the first forecast model；
Forecast model signal generating unit 220, for obtaining historical traffic flows data, historical traffic stream described in sliding-model control
Amount data, historical traffic flow data after sliding-model control is carried out the space time correlation rule that data mining obtains, and composition second is pre-
Survey model.
In this specification, each embodiment is described by the way of going forward one by one, and what each embodiment stressed is and other
The difference of embodiment, between each embodiment identical similar portion mutually referring to.For device disclosed in embodiment
For, because it corresponds to the method disclosed in Example, so description is fairly simple, say referring to method part in place of correlation
Bright.
Described above to the disclosed embodiments, makes professional and technical personnel in the field be capable of or uses the present invention.
Multiple modifications to these embodiments will be apparent from for those skilled in the art, as defined herein
General Principle can be realized without departing from the spirit or scope of the present invention in other embodiments.Therefore, the present invention
It is not intended to be limited to the embodiments shown herein, and be to fit to and principles disclosed herein and features of novelty phase one
The scope the widest causing.
Claims (8)
1. a kind of traffic flow forecasting method, comprising:
Obtain current traffic flow data, described current traffic flow data include section to be measured current traffic flow data and
Has the Current traffic flow data in relatedness section with section to be measured；
Determine the first forecast model；
By the first forecast model described in described current traffic flow data input, obtain the predicting traffic flow value in section to be measured；
Wherein, described first forecast model passes through to obtain historical traffic flows data, historical traffic flows number described in discrete processes
According to, obtain discretization historical data, described discretization historical data is carried out data mining obtain space time correlation rule, by described
Space time correlation rule obtains the section set having relatedness with section to be measured, and the historical traffic flow data of described section set is led to
Cross preordering method training to generate；
It is characterized in that, described determination the first forecast model includes:
Current traffic flow data described in discrete processes, obtains discretization current data；
Described discretization current data is inputted the second forecast model, obtains traffic flow trend；
According to corresponding relation between described section to be measured identification information, current time information and traffic flow trend and the first forecast model
Determine the first forecast model matching with the second forecast model；
Wherein, described second forecast model passes through to obtain historical traffic flows data, historical traffic flows described in sliding-model control
Data, historical traffic flow data after sliding-model control is carried out the space time correlation rule composition that data mining obtains.
2. traffic flow forecasting method according to claim 1 it is characterised in that
According to current traffic flow data described in following steps discrete processes:
The current traffic flow data normalization obtaining is interval interior to [0,1]；
Discretization is carried out to current traffic flow data after described normalized, obtains discretization current data；
According to historical traffic flows data described in following steps discrete processes:
The historical traffic circuit-switched data of acquisition is normalized to [0,1] interval interior；
Discretization is carried out to historical traffic flows data after described normalized, obtains discretization historical data.
3. traffic flow forecasting method according to claim 1 it is characterised in that by evolutionary programming algorithm to described from
Dispersion historical data carries out data mining and obtains space time correlation rule；
By evolutionary programming algorithm described discretization historical data is carried out data mining obtain space time correlation rule include:
Determine the number m in section in discretization historical data, generate the individuality with m gene；
Take the individual composition initial population not less than predetermined number, and calculated in described initial population according to discretization historical data
Each individual fitness value；
Swarm Evolution is carried out to described initial population according to each individual fitness value in described initial population；
Qualified correlation rules all in each individuality after Swarm Evolution are exported, is obtained space time correlation rule.
4. traffic flow forecasting method according to claim 3 is it is characterised in that described generation has the individual of m gene
Body includes:
Generate the determination gene that a span is [- t, -1], t is traffic flow data sampling in historical traffic flows data
The quantity of time interval；
Generate the random gene that m-1 span is [0, t], obtain the individuality with m gene.
5. traffic flow forecasting method according to claim 3 it is characterised in that described according to every in described initial population
Individual fitness value carries out Swarm Evolution to described initial population and includes:
Randomly choosing two individualities in described initial population, selecting the larger individual entrance of wherein fitness value of future generation, until selecting
Go out the new colony of the individual composition of predetermined number；
Randomly choose out predetermined number individuality and complete chromosomal chiasma operation from described new colony；
Randomly choose out predetermined number individuality and complete chromosome disorder operation from described new colony；
Repeat new colony and generate operation, chromosomal chiasma operation and chromosome disorder operation, until colony's fitness value is received
Hold back.
6. traffic flow forecasting method according to claim 1 is it is characterised in that described current traffic flow data and institute
State historical traffic flows data and be and filtered out abnormal data, and the abnormal data filtering out was carried out with the number revised
According to.
7. traffic flow forecasting method according to claim 1 is it is characterised in that section described and to be measured has relatedness road
The Current traffic stream packets of section include:
Determine section to be measured and space time correlation rule；
Obtain the section having relatedness with section to be measured according to described section to be measured and space time correlation rule.
8. a kind of traffic flow forecasting system, comprising: acquisition module, model determining module, prediction module and forecast model generate
Module；Wherein,
Described acquisition module, for obtaining current traffic flow data, described current traffic flow data includes section to be measured
Current traffic flow data and the Current traffic flow data having relatedness section with section to be measured；
Described model determining module, for determining the first forecast model；
Described prediction module, for by the first forecast model described in described current traffic flow data input, obtaining section to be measured
Predicting traffic flow value；
Described forecast model generation module, for obtaining historical traffic flows data, historical traffic flows described in sliding-model control
Data, historical traffic flow data after sliding-model control is carried out data mining and obtains space time correlation rule, by described space time correlation
Rule obtains the section set having relatedness with section to be measured, and the historical traffic flow data of described section set is passed through predetermined party
Method training generates the first forecast model；
It is characterized in that, described model determining module includes: forecast model determining unit and forecast model signal generating unit；Wherein,
Described forecast model determining unit, for current traffic flow data described in sliding-model control, obtains discretization current number
According to；Described discretization current data is inputted the second forecast model, obtains traffic flow trend；According to described traffic flow trend with
The first forecast model that between the first forecast model, corresponding relation determination is matched with the second forecast model；
Described forecast model signal generating unit, for obtaining historical traffic flows data, historical traffic flows described in sliding-model control
Data, historical traffic flow data after sliding-model control is carried out the space time correlation rule that data mining obtains, composition second prediction
Model.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410742725.8A CN104464291B (en) | 2014-12-08 | 2014-12-08 | Traffic flow predicting method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410742725.8A CN104464291B (en) | 2014-12-08 | 2014-12-08 | Traffic flow predicting method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104464291A CN104464291A (en) | 2015-03-25 |
| CN104464291B true CN104464291B (en) | 2017-02-01 |
Family
ID=52910266
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410742725.8A Active CN104464291B (en) | 2014-12-08 | 2014-12-08 | Traffic flow predicting method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104464291B (en) |
Families Citing this family (27)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105513350A (en) * | 2015-11-30 | 2016-04-20 | 华南理工大学 | Time-phased multi-parameter short-term traffic flow prediction method based on time-space characteristics |
| CN105788270A (en) * | 2016-05-13 | 2016-07-20 | 广州运星科技有限公司 | Internet of things-based traffic data prediction method and processing server |
| CN107423742A (en) * | 2016-05-23 | 2017-12-01 | 中兴通讯股份有限公司 | The determination method and device of crowd's flow |
| CN105957329B (en) * | 2016-06-29 | 2019-04-19 | 芜湖达成储运有限公司 | A kind of highway information intelligence system |
| CN106408450A (en) * | 2016-09-09 | 2017-02-15 | 国家电网公司 | Power distribution capability evaluating method |
| CN108230698B (en) * | 2016-12-21 | 2020-06-26 | 杭州海康威视数字技术股份有限公司 | Traffic flow prediction method and device |
| CN106875681B (en) * | 2017-03-22 | 2019-08-13 | 山东大学 | A kind of traffic congestion tendency prediction technique and system |
| CN107123267A (en) * | 2017-06-29 | 2017-09-01 | 中国路桥工程有限责任公司 | A kind of Freeway Traffic Volume Prediction system and method |
| CN107341252B (en) * | 2017-07-10 | 2018-08-17 | 北京神州泰岳软件股份有限公司 | A kind of method and device of the unknown incidence relation of mining rule correlation model |
| CN107688653B (en) * | 2017-09-01 | 2020-12-01 | 武汉倚天剑科技有限公司 | User behavior data mining system and method based on network shallow data |
| CN107943861A (en) * | 2017-11-09 | 2018-04-20 | 北京众荟信息技术股份有限公司 | A kind of missing data compensation process and system based on time series |
| CN108280542B (en) * | 2018-01-15 | 2021-05-11 | 深圳市和讯华谷信息技术有限公司 | User portrait model optimization method, medium and equipment |
| CN110098944B (en) * | 2018-01-29 | 2020-09-08 | 中国科学院声学研究所 | Method for predicting protocol data traffic based on FP-Growth and RNN |
| CN110444009B (en) * | 2018-05-02 | 2022-01-21 | 广州智丰设计研发有限公司 | Expressway traffic flow prediction system based on Internet of things |
| CN110444010B (en) * | 2018-05-02 | 2022-01-21 | 广州智丰设计研发有限公司 | Expressway traffic flow prediction method based on Internet of things |
| CN108682152B (en) * | 2018-05-31 | 2021-01-15 | 重庆大学 | Bus abnormal data restoration method based on preceding bus data |
| CN108564789B (en) * | 2018-06-11 | 2020-11-17 | 重庆大学 | Expressway ramp up-down direction flow estimation method based on toll station data |
| CN108898851B (en) * | 2018-06-20 | 2020-11-27 | 东南大学 | Combined prediction method for traffic volume of urban road section |
| CN109002988B (en) * | 2018-07-18 | 2023-10-27 | 平安科技（深圳）有限公司 | Risk passenger flow prediction method, apparatus, computer device and storage medium |
| CN109615852B (en) * | 2018-11-20 | 2021-03-30 | 上海海事大学 | Method for rapidly predicting road traffic flow distribution under sudden accident |
| CN110264725B (en) * | 2019-07-31 | 2021-08-03 | 北京世纪高通科技有限公司 | Method and device for determining road section flow |
| CN110444022A (en) * | 2019-08-15 | 2019-11-12 | 平安科技（深圳）有限公司 | The construction method and device of traffic flow data analysis model |
| CN110543978A (en) * | 2019-08-15 | 2019-12-06 | 平安科技（深圳）有限公司 | Traffic flow data prediction method and device based on wavelet neural network |
| CN111462484B (en) * | 2020-03-31 | 2021-11-30 | 腾讯科技（深圳）有限公司 | Congestion state determination method, device, equipment and computer readable storage medium |
| CN112561089A (en) * | 2020-11-27 | 2021-03-26 | 成都飞机工业（集团）有限责任公司 | Correlation analysis and prediction method for vulnerable spare parts |
| CN112598904A (en) * | 2020-12-10 | 2021-04-02 | 南通大学 | Urban road network traffic flow prediction method based on linear gray convolution model |
| CN113240917B (en) * | 2021-05-08 | 2022-11-08 | 广州隧华智慧交通科技有限公司 | Traffic management system applying deep neural network to intelligent traffic |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101739818A (en) * | 2009-11-19 | 2010-06-16 | 北京世纪高通科技有限公司 | Method and device for data mining of road traffic accident based on association rule |
| CN102708684B (en) * | 2012-06-21 | 2013-10-16 | 陕西师范大学 | Short-term traffic flow Volterra-DFP self-adaption prediction method |
| CN103116808A (en) * | 2013-01-18 | 2013-05-22 | 同济大学 | Method of real-timely predicting short time traffic flow of express way |
| CN103413443B (en) * | 2013-07-03 | 2015-05-20 | 太原理工大学 | Short-term traffic flow forecasting method based on hidden Markov model |
-
2014
- 2014-12-08 CN CN201410742725.8A patent/CN104464291B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN104464291A (en) | 2015-03-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104464291B (en) | Traffic flow predicting method and system | |
| Chen et al. | A short-term traffic prediction model in the vehicular cyber–physical systems | |
| CN102542818B (en) | A kind of coordination control method for traffic signal of zone boundary based on organic calculating | |
| Zhou et al. | Optimizing taxi driver profit efficiency: A spatial network-based markov decision process approach | |
| CN105163326B (en) | A kind of cell clustering method and system based on wireless network traffic feature | |
| CN102568205B (en) | Traffic parameter short-time prediction method based on empirical mode decomposition and classification combination prediction in abnormal state | |
| CN102469103B (en) | Trojan event prediction method based on BP (Back Propagation) neural network | |
| CN106022614A (en) | Data mining method of neural network based on nearest neighbor clustering | |
| CN104883278A (en) | Method for classifying network equipment by utilizing machine learning | |
| CN102346964A (en) | Real-time jam prediction and intelligent management system for road traffic network area | |
| CN109376906B (en) | Travel time prediction method and system based on multi-dimensional trajectory and electronic equipment | |
| CN109215344B (en) | Method and system for urban road short-time traffic flow prediction | |
| CN106067079A (en) | A kind of system and method for gray haze based on BP neutral net prediction | |
| CN114553661A (en) | Mobile user equipment clustering training method for wireless federal learning | |
| CN109978215A (en) | Patrol management method and device | |
| CN111737826B (en) | Rail transit automatic simulation modeling method and device based on reinforcement learning | |
| CN107481523A (en) | A kind of traffic flow speed Forecasting Methodology and system | |
| Massoud et al. | Exploring fuzzy logic and random forest for car drivers’ fuel consumption estimation in IoT-enabled serious games | |
| CN104217296A (en) | Listed company performance comprehensive evaluation method | |
| CN113688200B (en) | Decision tree-based special population action track collection method and system | |
| CN114418177B (en) | New product material distribution prediction method based on digital twin workshops for generating countermeasure network | |
| CN115691140A (en) | Analysis and prediction method for space-time distribution of automobile charging demand | |
| CN114841451A (en) | Driver travel subsidy method and device and storage medium | |
| CN114331188A (en) | Project value evaluation reference method and system based on big data | |
| Ye et al. | Hybrid calibration of agent-based travel model using traffic counts and AVI data |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant |