CN106781508B - Short-time Traffic Flow Forecasting Methods based on multiple phase space under a kind of Spark environment - Google Patents
Short-time Traffic Flow Forecasting Methods based on multiple phase space under a kind of Spark environment Download PDFInfo
- Publication number
- CN106781508B CN106781508B CN201710110044.3A CN201710110044A CN106781508B CN 106781508 B CN106781508 B CN 106781508B CN 201710110044 A CN201710110044 A CN 201710110044A CN 106781508 B CN106781508 B CN 106781508B
- Authority
- CN
- China
- Prior art keywords
- traffic flow
- time
- vehicle
- data
- phase space
- 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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/21—Design, administration or maintenance of databases
- G06F16/215—Improving data quality; Data cleansing, e.g. de-duplication, removing invalid entries or correcting typographical errors
Landscapes
- Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Quality & Reliability (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Traffic Control Systems (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses the Short-time Traffic Flow Forecasting Methods based on multiple phase space under a kind of Spark environment, comprising the following steps: step (1): obtaining the real-time vehicle information in some section, and is sent in database HBase；Step (2): it according to the history information of vehicles stored in database HBase, obtains historical traffic flow data and stores in the database；Step (3): under Spark environment, building multiplephase spatial model is predicted in conjunction with historical traffic flow data with traffic flow data of the model to next period.Using technical solution of the present invention, multiple phase space and big data frame Spark technology, the more scientific accurately traffic flow of prediction subsequent time in real time are constructed by multipair delay time and insertion dimension.
Description
Technical field
The present invention relates to intelligent transport system field, in particular to based on the short of multiple phase space under a kind of Spark environment
Phase traffic real-time predicting method.
Background technique
With the fast development of Chinese economy, transportation is while bringing huge economy and society effect,
It is faced with huge government pressure.The vehicle of huge amount causes traffic congestion, aggravates air pollution, slows down traffic efficiency, seriously
Ground affects the trip of people.Forecasting traffic flow timely accurately is carried out using telecommunication flow information, manager can be helped to make
Reasonable traffic control scheme, provides huge reference value for the trip decision-making of people.Traffic flow systems are one very multiple
Miscellaneous nonlinear system, and meet chaology characteristic.Chaology is as change over time one of research nonlinear system
Science can preferably react the inherent law of Traffic flow systems.The traffic forecast method of the prior art is primarily present following two
A problem:
1) prediction of short-term traffic volume model is all predicted under stand-alone environment, and traditional database is not able to satisfy the height of traffic data
Effect storage calculates, and reaches time delay, brings challenge for the data processing of prediction model.Nowadays, we have come into friendship
Logical big data era, data information become increasingly abundant, and stand-alone environment is difficult to handle the big data of magnanimity.Spark makees in the prior art
Frame is handled for the big data of a very outstanding open source, referring to Fig. 1, the functional block diagram of shown Spark framework, the frame base
Data processing speed is improved in memory calculating, and bottom uses high-performance distributed PostgreSQL database HBase, greatly improve
Data storage capacities have obtained being widely applied very much in various fields.
2) traditional algorithm mainly utilizes shallow-layer traffic prediction model, but it does not consider the chaotic characteristic of traffic system,
Uncertain strong short-time traffic flow forecast is not adapted to.And though traditional chaos algorithm considers chaotic characteristic, using individually prolonging
The slow time is predicted that precision of prediction is unsatisfactory with insertion dimension.Therefore, there is an urgent need to a kind of in real time and accurate pre- at present
The method for surveying road Short-term Traffic Flow, makes people more freely select the travel time, farthest to guarantee congestion problems
Slow down, so that road traffic problem is further improved.
Therefore for drawbacks described above present in currently available technology, it is really necessary to be studied, to provide a kind of scheme,
Solve defect existing in the prior art.
Summary of the invention
In view of this, being predicted in real time it is necessory to provide the short-term traffic based on multiple phase space under a kind of Spark environment
Method, this method are using big data Spark technology, and to establish multiplephase under the traffic flow data information environment of magnanimity
Spatial prediction model, so that the traffic flow data to subsequent period is accurately predicted in real time.
In order to solve problems in the prior art, technical scheme is as follows:
Short-time Traffic Flow Forecasting Methods based on multiple phase space under a kind of Spark environment, comprising the following steps:
Step (1): the real-time vehicle information in some section is obtained, and is sent in database HBase；
Step (2): it according to the history information of vehicles stored in database HBase, obtains historical traffic flow data and stores
In the database；
Step (3): under Spark environment, multiplephase spatial model is constructed, in conjunction with historical traffic flow data, with the mould
Type predicts the traffic flow data of next period；
Wherein, the step (3) further includes steps of
Step (31): to Mr. Yu section R, delay time set S is determinedτ, it is embedded in dimension collection SmAnd corresponding weight
Distribution
Step (32): according to insertion dimension miPredicting traffic flow time series X is determined, further according to delay time TiBefore obtaining
Set time series traffic flow vector Xf；
Step (33): taking the historical traffic flow data in the R of section, successively constitutes time series traffic flow vector X1,X2…
Xn, calculating and XfSimilarity and ascending order arrange d1,d2…dn；
Step (34): it is found from historical traffic flow data and " lead time sequence " most similar z " correlation time sequences
Column ", that is, z distance d before obtaining1,d2…dz, corresponding vector is X1,X2…Xz, respective weights distributionS=1, dm=min { d1,d2…dz}；
Step (35): by time series traffic flow vectorAnd weight distributionIn conjunction with delay time weight set Q τ,
According to formulaWeighted calculation obtains part traffic flow；
Step (36): repeating step (32) to step (35), and taking every a pair of { τ, m } is one group of phase space, and substitution calculates
To part traffic flow vectorThese results are added up and obtain final predicting traffic flow Xv:
Preferably, it in the step (1), further includes steps of
It is obtained using road traffic monitoring equipment and image processing algorithm through vehicle license plate information, and is sent to database
In；
Maintenance data cleans algorithm and carries out depth cleaning to the vehicle data in database, rejects invalid data information.
Preferably, the step (2) further includes steps of
Traffic information set is established using following formula:
Wherein, C indicates that vehicle collects information aggregate, HiIndicating the license plate number of i-th vehicle, N is total vehicle number,
Bi,jIndicate that the j-th strip bayonet of i-th vehicle crosses vehicle and records corresponding bayonet number, Ti,jIndicate that the j-th strip bayonet of i-th vehicle crosses vehicle
Record corresponding vehicle time excessively, TlowIndicate lower limit, the T of setting timeupIndicate the upper limit of setting time, MiIndicate HiNumber vehicle exists
Vehicle data sum in set period of time；
The vehicle fleet between bayonet is calculated by following algorithm:
Wherein, l () represents Boolean expression, l (true)=1, l (false)=0；
To the information of vehicles set in all sections in the traffic flow data of magnanimityDuplicate execution above-mentioned steps obtain
Vehicle fleet P of certain a road section in certain a period of timen,m, time interval t calculates the friendship in each section by following formula
It is through-flow:
Fn,m=Pn,m/t。
Compared with prior art, the invention has the benefit that
1. accuracy: the present invention constructs multiple phase space using multipair delay time and insertion dimension, forms prediction model,
The traffic flow of more scientific accurate prediction subsequent time.
2. real-time: the present invention uses big data frame Spark technology, faces large-scale historical sample data, helps
In quick calculating and storing data, be conducive to the real-time prediction of subsequent time traffic flow.
Detailed description of the invention
Fig. 1 is that Spark big data handles frame architecture diagram；
Fig. 2 is the overview flow chart of road traffic flow method of real-time；
Fig. 3 is to construct algorithm under multiple phase space to predict overview flow chart；
Fig. 4 is the curve graph of Average I (τ) about delay time T；
Fig. 5 is " volume of traffic-time " figure of Wenzhou City's Western Hills East Road forecasting traffic flow.
Following specific embodiment will further illustrate the present invention in conjunction with above-mentioned attached drawing.
Specific embodiment
It is pre- in real time to the short-term traffic based on multiple phase space under Spark environment provided by the invention below with reference to attached drawing
Survey method is described further.
The central scope of technical solution of the present invention is: the real-time vehicle data at some section each moment are stored in data
In the HBase of library, and then the traffic flow data in available any time period.Under Spark environment, multiplephase spatial mode is constructed
Type uses multiplephase spatial model to next by the traffic flow data of the traffic flow data combination historical period of present period
The traffic flow data of period is predicted.To realize the forecast of urban traffic based on Current traffic flow data.Tool
It says to body, in order to predict certain a road section in the magnitude of traffic flow of certain following period, we are predicted first comprising the magnitude of traffic flow
Time series (" predicted time sequence "), obtains the magnitude of traffic flow from the time series in advance.For every a pair of of delay time T and
It is embedded in dimension m, obtains " the lead time sequence " of " predicted time sequence ", is found in historical traffic flow data and " when preposition
Between sequence " most similar z " time related sequence ", obtain " predicted time sequence based on this z " time related sequence " predictions
Column ".In this way for every a pair of of delay time T and insertion dimension m, we have obtained the magnitude of traffic flow, finally pre- measure these
To the magnitude of traffic flow integrate the magnitude of traffic flow as the section.
Referring to fig. 2, it is shown the step flow chart for obtaining traffic flow in the method for the present invention, is specifically comprised the following steps:
Step (S21) is obtained by vehicle license plate information, concurrently using road traffic monitoring equipment and image processing algorithm
It send into central database.
Step (S22) maintenance data cleans relevant algorithm (exceptional value cleaning, missing values cleaning etc.) in database
Vehicle data carries out depth cleaning, rejects invalid data information (vehicle is unidentified, and license plate number is invalid etc.).
Step (S23) is analyzed and processed the information of vehicles of above-mentioned acquisition and calculates traffic flow；Assuming that one section of road is deposited
In a pair of of bayonet Bn, Bm, for some period Ti, Tj, the period meets Tlow≤Ti≤Tj≤Tup, setting time difference threshold value
εh, when | Tj-Ti|<εh, vehicle number Pn,mIncrease, it is invalid to be otherwise regarded as stopping.Calculate traffic flow Fn,m=Pn,m/(Tj-Ti)。
Wherein, it is further included steps of in step (S23)
Step (S231): bayonet cross vehicle be recorded as<bayonet number, license plate number, cross the vehicle time>.It is established using following formula
Traffic information set:
C indicates that vehicle collects information aggregate, HiIndicate the license plate number of i-th vehicle, N is total vehicle number, Bi,jIt indicates
The j-th strip bayonet of i-th vehicle crosses vehicle and records corresponding bayonet number, Ti,jIndicate that the j-th strip bayonet of i-th vehicle crosses vehicle record pair
That answers spends the vehicle time, TlowIndicate lower limit, the T of setting timeupIndicate the upper limit of setting time, MiIndicate HiNumber vehicle is in setting
Between vehicle data sum in section；
Step (S232): the vehicle fleet between bayonet is calculated by following algorithm:
L () represents Boolean expression, l (true)=1, l (false)=0.
Step (S233): to the information of vehicles set in all sections in the traffic flow data of magnanimityIt is duplicate to execute step
Rapid step (S232), obtains certain a road section in the vehicle fleet P of certain a period of timen,m, time interval t, by following formula come
Calculate the traffic flow in each section.
Fn,m=Pn,m/t
Referring to Fig. 3, it show in the method for the present invention using the specific flow chart of multiplephase spatial model prediction, further wraps
Include following steps:
Step (S31) determines delay time set Sτ, it is embedded in dimension collection Sm, multiple phase space is collectively constituted, and corresponding
Weight distribution
Wherein, step (S31) further includes steps of
Step (S311): pass through mutual information method computing relay time set Sτ={ τ1,τ2...τnτ, when each postpones
Between the corresponding Average I (τ) of τ, draw to obtain Average I (τ) about delay time T by experimental calculation
Curve graph, such as Fig. 4.The following table 1 shows the part minimum point in curve graph.
The part minimum point of table 1 τ and I (τ) curve graph
As shown in Table 1, first minimum point is τ=7, I=3.311, therefore the minimum point is optimum delay time
τ.Later, this paper subsequently selected minimum point is necessarily less than its previous minimum point, and then second minimum point is τ
=37, I=3.305 repeat this process.The optimum delay time set S finally selectedτ={ 7,37,45,61 }, according to such as
Lower formula carries out weight distribution:Delay time weight set Qτ=0.68,0.13,0.11,
0.08}。
Step (S312): the Joint Distribution for setting two stochastic variables (X, Y) is p (x, y), and limit distribution is respectively p (x), p
(y), mutual information I (X；Y) be Joint Distribution p (x, y) Yu product distribution p (x) p (y) relative entropy, i.e.,
Step (S313): Grassberger and proeaeeia is according to the thought of embedding theory and phase space reconstruction, first
Propose the method that insertion dimension is directly calculated from data sequence, commonly referred to as G-P method.Insertion dimension is directly calculated by G-P method
M is spent, according to some delay time and known Traffic Flow Time Series, constructs the multi-dimensional state vector that an insertion dimension is m
Embedded space,Wherein,Indicate section i in time period tj
The interior magnitude of traffic flow, p are total period number.The vector distance between two vectors is solved, as long as and stipulated that distance
Less than the vector of given positive number ε, become the vector that is mutually related, the ratio that these vectors that are mutually related account for total vector is made
The correlation integral of phase space is tieed up for reconstruct m:Wherein, H is He Wei Saden rank
Jump function, di,jIt is Euler's normal form, it is vector XiAnd XjThe distance between, indicate similarity.Finally define correlation dimension D2,The correlation dimension D of chaos time sequence2It can be with insertion dimension m
Increase and restrains, the correlation dimension D of immiscible chaos time series2It can be discrete with insertion dimension m increase.Therefore available insertion dimension
The convergence of m is spent to judge whether the magnitude of traffic flow meets chaology.According to the delay time set S of step (S311)τ, it is every
Insertion dimension m (τ) is calculated in one delay time T, obtains insertion dimension collection Sm。
Step (S32): according to insertion dimension miPredicting traffic flow time series X is determined, further according to delay time TiBefore obtaining
Set time series traffic flow vector Xf。
Step (S33): finding and XfMost similar z historical traffic stream time series.
The historical traffic flow data in the section is taken, time series traffic flow vector X is successively constituted1,X2…Xn, calculate and Xf
Similarity and ascending order arrangement.
Step (S34): z distance d before obtaining1,d2…dz, corresponding vector is X1,X2…Xz, form weight distributionS=1, dm=min { d1,d2…dz}。
Step (S35): by time series traffic flow vectorWeight distributionIn conjunction with delay time weight set Qτ,
According to formulaWeighted calculation obtains part traffic flow.
Step (S36): repeating step (S32) to step (S35), takes every a pair of { τ, m } to substitute into and part traffic is calculated
Flow vector obtains final predicting traffic flow Xv:
It is described in detail by taking the Shandong road of Wenzhou City road network Chinese and Western as an example below:
We use the GPS data that on March 25th, 2015 to March 27, totally three days entire Wenzhou City's road networks generated.Pass through
Road monitoring equipment obtains section information of vehicles in real time, obtains the specifying information of vehicle with image processing algorithm and is stored in number
According to library, obtain by vehicular traffic number Pn,m, and pass through formula Fn,m=Pn,m/ t calculates traffic flow and is stored in database profession.
Delay time set S is obtained with related algorithmτ={ 7,37,45,61 } are embedded in dimension collection Sm={ 5,8,10,11 }, weight
Set Qτ={ 0.65,0.14,0.12,0.09 } will predict certain bayonet to the traffic flow F in 8:00~8:108:00~8:10, first
Take τ=7, m=5.According to m=5, predicting traffic flow time series X={ F7:20~7:30,F7:30~7:40…F8:00~8:10, according to τ=
7, then preposition Traffic Flow Time Series Xf={ F6:10~6:20,F6:20~6:30…F6:50~7:00, it is found from historical traffic stream sequence
With XfMost similar z time series, and thus prediction obtains a traffic flow F8:00~8:10Value.Each pair of delay time and insertion
Dimension substitutes into prediction and obtains a traffic flow F8:00~8:10Value obtains final F by weighted average8:00~8:10Value.
Fig. 5 illustrates this paper algorithm under March 27, and the effect between prediction result and traditional algorithm compares, when prediction
Between be 6:00~18:00, this paper algorithm predicted value and measured value accuracy are very high, traditional algorithm difference it is larger, and data shakiness
It is fixed.
In conclusion the present invention is stored in database in real time after pretreatment by real-time monitoring traffic data, pass through this
The formula F of methodn,m=Pn,m/ t calculates the traffic flow of every road and is stored in HBase, by multiplephase space arithmetic mould
Type is predicted to obtain the traffic flow of certain section subsequent period.This method has the characteristics that accuracy and real-time, overcomes traditional
Traffic prediction model is predicted to obtain deficiency of the method for road traffic flow in terms of real-time, accuracy under big data background.
The above description of the embodiment is only used to help understand the method for the present invention and its core ideas.It should be pointed out that pair
For those skilled in the art, without departing from the principle of the present invention, the present invention can also be carried out
Some improvements and modifications, these improvements and modifications also fall within the scope of protection of the claims of the present invention.
The foregoing description of the disclosed embodiments enables those skilled in the art to implement or use the present invention.
Various modifications to these embodiments will be readily apparent to those skilled in the art, as defined herein
General Principle can be realized in other embodiments without departing from the spirit or scope of the present invention.Therefore, of the invention
It is not intended to be limited to the embodiments shown herein, and is to fit to and the principles and novel features disclosed herein phase one
The widest scope of cause.
Claims (3)
1. the Short-time Traffic Flow Forecasting Methods based on multiple phase space under a kind of Spark environment, which is characterized in that including following step
It is rapid:
Step (1): the real-time vehicle information in some section is obtained, and is sent in database HBase；
Step (2): it according to the history information of vehicles stored in database HBase, obtains historical traffic flow data and is stored in number
According in library；
Step (3): under Spark environment, multiplephase spatial model is constructed, in conjunction with historical traffic flow data, with the model pair
The traffic flow data of next period is predicted；
Wherein, the step (3) further includes steps of
Step (31): to Mr. Yu section R, delay time set S is determinedτ, it is embedded in dimension collection Sm, and
Corresponding weight distribution
Step (32): according to insertion dimension miPredicting traffic flow time series X is determined, further according to delay time TiWhen obtaining preposition
Between sequence traffic flow vector Xf；
Step (33): taking the historical traffic flow data in the R of section, successively constitutes time series traffic flow vector X1,X2…Xn, meter
Calculation and XfSimilarity and ascending order arrange d1,d2…dn；
Step (34): finding from historical traffic flow data and " lead time sequence " most similar z is a " time related sequence ",
Z distance d before obtaining1,d2…dz, corresponding vector is X1,X2…Xz, respective weights distributionS=1, dm=min { d1,d2…dz}；
Step (35): by time series traffic flow vectorAnd weight distributionIn conjunction with delay time weight set Qτ, according to
FormulaWeighted calculation obtains part traffic flow；
Step (36): repeating step (32) to step (35), and taking every a pair of { τ, m } is one group of phase space, and portion is calculated in substitution
Divide traffic flow vectorThese results are added up and obtain final predicting traffic flow Xv:
2. the Short-time Traffic Flow Forecasting Methods based on multiple phase space under Spark environment according to claim 1, feature
It is, in the step (1), further includes steps of
It is obtained using road traffic monitoring equipment and image processing algorithm through vehicle license plate information, and is sent in database；
Maintenance data cleans algorithm and carries out depth cleaning to the vehicle data in database, rejects invalid data information.
3. the Short-time Traffic Flow Forecasting Methods based on multiple phase space under Spark environment according to claim 1 or 2,
It is characterized in that, the step (2) further includes steps of
Traffic information set is established using following formula:
Wherein, C indicates that vehicle collects information aggregate, HiIndicate the license plate number of i-th vehicle, N is total vehicle number, Bi,jIt indicates
The j-th strip bayonet of i-th vehicle crosses vehicle and records corresponding bayonet number, Ti,jIndicate that the j-th strip bayonet of i-th vehicle crosses vehicle record pair
That answers spends the vehicle time, TlowIndicate lower limit, the T of setting timeupIndicate the upper limit of setting time, MiIndicate HiNumber vehicle is in setting
Between vehicle fleet in section；
The vehicle fleet between bayonet is calculated by following algorithm:
Wherein, l () represents Boolean expression, l (true)=1, l (false)=0；
To the information of vehicles set C in all sections in the traffic flow data of magnanimityHi, duplicate execution above-mentioned steps obtain certain all the way
Vehicle fleet P of the section in certain a period of timen,m, time interval t calculates the traffic flow in each section by following formula:
Fn,m=Pn,m/t。
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710110044.3A CN106781508B (en) | 2017-02-28 | 2017-02-28 | Short-time Traffic Flow Forecasting Methods based on multiple phase space under a kind of Spark environment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710110044.3A CN106781508B (en) | 2017-02-28 | 2017-02-28 | Short-time Traffic Flow Forecasting Methods based on multiple phase space under a kind of Spark environment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106781508A CN106781508A (en) | 2017-05-31 |
| CN106781508B true CN106781508B (en) | 2019-01-25 |
Family
ID=58960045
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710110044.3A Active CN106781508B (en) | 2017-02-28 | 2017-02-28 | Short-time Traffic Flow Forecasting Methods based on multiple phase space under a kind of Spark environment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106781508B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108281013A (en) * | 2018-03-22 | 2018-07-13 | 安徽八六物联科技有限公司 | A kind of road traffic monitoring system |
| CN111063189B (en) * | 2019-11-05 | 2022-05-03 | 阿里巴巴集团控股有限公司 | Traffic flow processing method and device and electronic equipment |
| CN112967495B (en) * | 2021-01-29 | 2023-05-09 | 贵州民族大学 | Short-time traffic flow prediction method and system based on big data of movement track |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105046953A (en) * | 2015-06-18 | 2015-11-11 | 南京信息工程大学 | Short-time traffic-flow combination prediction method |
| CN105303835A (en) * | 2015-11-13 | 2016-02-03 | 西安邮电大学 | Short-time prediction method of road traffic flow state |
| CN105701571A (en) * | 2016-01-13 | 2016-06-22 | 南京邮电大学 | Short-term traffic flow prediction method based on nerve network combination model |
| 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 |
| WO2016187129A1 (en) * | 2015-05-20 | 2016-11-24 | Continental Automotive Systems, Inc. | Generating predictive information associated with vehicle products/services |
-
2017
- 2017-02-28 CN CN201710110044.3A patent/CN106781508B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2016187129A1 (en) * | 2015-05-20 | 2016-11-24 | Continental Automotive Systems, Inc. | Generating predictive information associated with vehicle products/services |
| CN105046953A (en) * | 2015-06-18 | 2015-11-11 | 南京信息工程大学 | Short-time traffic-flow combination prediction method |
| CN105303835A (en) * | 2015-11-13 | 2016-02-03 | 西安邮电大学 | Short-time prediction method of road traffic flow state |
| CN105701571A (en) * | 2016-01-13 | 2016-06-22 | 南京邮电大学 | Short-term traffic flow prediction method based on nerve network combination model |
| 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 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106781508A (en) | 2017-05-31 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109285346B (en) | Urban road network traffic state prediction method based on key road sections | |
| Fang et al. | FTPG: A fine-grained traffic prediction method with graph attention network using big trace data | |
| CN107610021B (en) | Comprehensive analysis method for space-time distribution of environment variables | |
| CN110570651A (en) | Road network traffic situation prediction method and system based on deep learning | |
| CN110517482B (en) | Short-term traffic flow prediction method based on 3D convolutional neural network | |
| CN104183119B (en) | Based on the anti-arithmetic for real-time traffic flow distribution forecasting method pushed away of section OD | |
| CN108898829A (en) | The dynamic short-time traffic flow forecast system with Sparse is divided for the indifference opposite sex | |
| CN109584552A (en) | A kind of public transport arrival time prediction technique based on network vector autoregression model | |
| CN110555476B (en) | Intelligent vehicle lane change track prediction method suitable for man-machine hybrid driving environment | |
| CN110299011A (en) | A kind of traffic flow forecasting method of the highway arbitrary cross-section based on charge data | |
| CN106781508B (en) | Short-time Traffic Flow Forecasting Methods based on multiple phase space under a kind of Spark environment | |
| CN112489426A (en) | Urban traffic flow space-time prediction scheme based on graph convolution neural network | |
| CN109934403A (en) | Charge load Analysis prediction technique in electric car resident region based on mathematical model | |
| CN106504535A (en) | A kind of combination Gravity Models and the trip distribution modeling method of Fratar models | |
| CN112633602B (en) | Traffic congestion index prediction method and device based on GIS map information | |
| CN106887141A (en) | Continuous Traffic node congestion degree forecast model based on queueing theory, system and method | |
| Yao et al. | Sampled trajectory data-driven method of cycle-based volume estimation for signalized intersections by hybridizing shockwave theory and probability distribution | |
| CN113762338A (en) | Traffic flow prediction method, equipment and medium based on multi-graph attention mechanism | |
| CN108549955A (en) | A kind of charging pile abnormal rate determines method and device | |
| CN103870890A (en) | Prediction method for traffic flow distribution of expressway network | |
| CN109064750A (en) | Urban road network traffic estimation method and system | |
| Zhou et al. | Method for judging parking status based on yolov2 target detection algorithm | |
| CN111784027A (en) | Urban range electric vehicle charging demand prediction method considering geographic information | |
| Zhang et al. | Off-deployment traffic estimation—a traffic generative adversarial networks approach | |
| Panovski et al. | Real-time public transportation prediction with machine learning algorithms |
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 | ||
| EE01 | Entry into force of recordation of patent licensing contract | ||
| EE01 | Entry into force of recordation of patent licensing contract |
Application publication date: 20170531 Assignee: Hangzhou Cheng Road Polytron Technologies Inc Assignor: Hangzhou Electronic Science and Technology Univ Contract record no.: 2019330000034 Denomination of invention: Short-time traffic flow prediction method based on multiphase space in Spark environment Granted publication date: 20190125 License type: Common License Record date: 20190319 |
|
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20220119 Address after: 310000 room 519, No. 78, Fengqi Road, Xiacheng District, Hangzhou City, Zhejiang Province Patentee after: HANGZHOU CHENGDAO TECHNOLOGY CO.,LTD. Address before: 310018 Xiasha Higher Education Zone, Hangzhou, Zhejiang Patentee before: HANGZHOU DIANZI University |