CN107170234A - A kind of method predicted based on SVM algorithm traffic in short-term - Google Patents
A kind of method predicted based on SVM algorithm traffic in short-term Download PDFInfo
- Publication number
- CN107170234A CN107170234A CN201710433951.1A CN201710433951A CN107170234A CN 107170234 A CN107170234 A CN 107170234A CN 201710433951 A CN201710433951 A CN 201710433951A CN 107170234 A CN107170234 A CN 107170234A
- Authority
- CN
- China
- Prior art keywords
- svm
- mrow
- road conditions
- traffic
- function
- 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.)
- Pending
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
- G08G1/0133—Traffic data processing for classifying traffic situation
-
- 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"
-
- 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
Landscapes
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Strategic Management (AREA)
- Human Resources & Organizations (AREA)
- Analytical Chemistry (AREA)
- Economics (AREA)
- Chemical & Material Sciences (AREA)
- Entrepreneurship & Innovation (AREA)
- Game Theory and Decision Science (AREA)
- Development Economics (AREA)
- Marketing (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Traffic Control Systems (AREA)
Abstract
The present invention provides a kind of method predicted based on SVM algorithm traffic in short-term, methods described is predicted road conditions value by SVM nonlinear regression, go to train SVM models according to the road conditions value before adjacent to the period, and then predict the road conditions value of next period, and the prediction that jam level completes each road traffic congestion situation in short-term is converted into according to road conditions value, go to obtain SVM models present invention employs the method for dynamically-adjusting parameter, determine kernel function optimized parameter, so as to reach raising precision of prediction, reduce the effect of error；Nonlinear regression based on SVM is predicted road conditions value, goes to train SVM models according to the road conditions value before adjacent to the period, and then predicts the road conditions value of next period, and is converted into the prediction that jam level completes each road traffic congestion situation in short-term according to road conditions value.
Description
Technical field
The invention belongs to technical field of intelligent traffic, and in particular to one kind is predicted traffic in short-term based on SVM algorithm
Method.
Background technology
Road traffic system is numerous people's participations, real-time change, complicated nonlinear system, not true with height
It is qualitative, and randomness.These factors all bring difficulty to traffic forecast, and especially prediction of short-term traffic volume is by random factors
Such as traffic accident, road construction, accident, Changes in weather influence are bigger, and uncertain bigger regularity is not more obvious.
Traffic flow has the complexity of height, non-linear and uncertainty, therefore the traffic set up based on classical mathematical method
Road condition predicting model, its precision of prediction is difficult to meet the demand that Real time Adaptive Traffic Control is induced in intelligent transportation system well.With
Going deep into for traffic forecast area research, various methods can substantially be divided into two classes：One class be using the traditional mathematicses such as mathematical statistics as
The Forecasting Methodology on basis；Another kind of is not pursue strict mathematical derivation, is more paid attention to the pre- of true traffic behavior fitting effect
Survey model.The first kind includes time series models, Kalman filter model, Partial Linear Models etc.；Equations of The Second Kind then includes non-ginseng
Number regression model, the method based on wavelet theory, neural network model etc..
Li Cunjun, Yang Rugui, Zhang Jiashu of Computers and Communication engineering college of Southwest Jiaotong University were being calculated in 2003
Paper has been delivered on machine application publication《Traffic flow forecasting method based on wavelet analysis》.In order to more accurately predict in paper
The magnitude of traffic flow of dynamic change, it is proposed that the method being predicted on the basis of wavelet analysis using Kalman Filtering for Discrete,
This method can be used for the different field that dynamic data is predicted, the prediction of such as network traffics, the prediction of economic information and other
The prediction of nonlinear system.Experiment shows that this method can effectively reduce the error of data prediction.Xinan Science and Technology Univ. believes
Cease engineering college Shen is intelligent, Liu Zhigui, Li Chunju have delivered paper in 2008 on Xinan Science and Technology Univ.'s journal《Based on BP god
Traffic flow forecasting design through network》.Using magnitude of traffic flow control as target in paper, in the Research foundation of traffic stream characteristics
On, the model of traffic flux forecast based on BP neural network is established, and verified by taking certain junctions section as an example, according to phase
The magnitude of traffic flow in adjacent two sections and weather conditions are modeled as input, as a result show that forecasting system can be predicted relatively accurately separately
The magnitude of traffic flow in one section.Swallow, Tang Shuming are known in 2003 in Chinese Highway journal in the palace of Institute of Automation Research of CAS
On delivered paper《Short-term traffic flow forecast and event detection integration algorithm based on non parametric regression》.Nonparametric in paper
Real Time Observation data are passed through k nearest neighbor searching algorithm by the main foundation cried according to historical traffic data of the forecasting traffic flow of recurrence
Matched, finally predicted the outcome in weighted average with historical data.This method is without training, and convenient transplanting, prediction is missed
Difference is smaller.
The existing method for traffic forecast is numerous, each advantageous and inferior position：Time series models are in a large amount of uninterrupted numbers
Precision is higher on the basis of, but parameter Estimation is complicated, and parameter can not be transplanted, and can cause in practical application because data omit problem
Precision of prediction is reduced, and relies on a large amount of historical datas, and cost is very high.The precision of prediction of Kalman filtering method is with predicted time interval
Change less, but every time calculate be both needed to adjust weights, calculate it is complex, it is difficult to used time real-time online predict, output result
Several periods can be postponed.Non parametric regression does not need priori, only needs historical data, by finding in historical data with working as
The similar neighbour of preceding point, with obtained neighbor prediction subsequent time result.This method is easy to use, and error distribution situation is good,
But the complexity of neighbour is searched in a large amount of historical datas may influence the ageing of output that predict the outcome.Neutral net relies on
The advantage such as the ability of its Approximation of Arbitrary Nonlinear Function and the fault-tolerant, self study that has, is used for by lot of domestic and international scholar
Model of traffic flux forecast is set up, and achieves many achievements in research.Because neutral net is a kind of the heuristic of dependence experience
Technology, his learning process uses empirical risk minimization principle, and under Small Sample Size, over-fitting easily occur causes
Generalization ability is low.Neural network algorithm complexity is influenceed larger by the complicated network structure and sample complexity simultaneously.This
It is a little not enough, make application effect of the neural network model in forecasting traffic flow desirably not good, for the short of non-stationary
When traffic flow, when input data is mixed with noise, Prediction Accuracy can be worse.
Therefore based on the analysis of above distinct methods, and the practical application scene in intelligent transportation system, it is determined that from being based on
The non-linear regression method of SVMs is predicted to carry out traffic in short-term.And the nonlinear regression based on SVMs
Prediction has strict theory and Fundamentals of Mathematics, and based on structural risk minimization, generalization ability is better than neutral net etc., algorithm
With Global Optimality.The juche idea of this method is, according to some training samples, to find an optimal function so that function
The expected risk (can temporarily be interpreted as error) between estimation Y' and reality output Y to inputting X is minimized.
The content of the invention
In order to solve the above problems, the present invention provides a kind of method predicted based on SVM algorithm traffic in short-term, institute
State method and road conditions value is predicted by SVM nonlinear regression, go to train SVM moulds according to the road conditions value before adjacent to the period
Type, and then the road conditions value of next period is predicted, and jam level completion each road traffic congestion in short-term is converted into according to road conditions value
The prediction of situation；
Further, methods described includes：
S1：Kernel function is selected, SVM parameters are determined, after the data set of road conditions value of neighbouring period is obtained, Gaussian kernel is selected
Function includes width parameter δ, the Optimal Parameters ε and C of quadratic programming as kernel function, the SVM parameters；
S2:The road conditions Value Data collection of input neighbouring period generates anticipation function as sample；
S3:Output result after evaluation analysis is carried out according to predicting the outcome；
S4:Traffic information according to representated by output result, carries out traffic in short-term and predicts；
Further, kernel function is selected in the S1, determines that the specific method of SVM parameters is as follows；
1) assume that training dataset is designated as T={ (xi,yi)}l I=1, first with Nonlinear Mapping ψ (x)=[ψ1(x),ψ2
(x),…,ψN(x)]TInput data from former space reflection to N-dimensional feature space, approximating function is constructed in feature space；
2) nonlinear regression of lower dimensional space is correspond in the linear regression of higher dimensional space, definition ε is insensitive loss letter
Number Lε(x, y, f)=| y-f (x) |ε=max (0, | y-f (x) |-ε)；
3) object function is minimized
Wherein, xi∈RnIt is i-th of input, yi∈ R are corresponding desired outputs；ω=[ω1, ω2..., ωn]TIt is line
Property weights variable, b for biasing；ε and C are two free parameters determined by user；
4) test sample x is predicted as the following formula, and correspondence is exported,
Wherein, K is the kernel function for meeting Mercer conditions, and the kernel function includes Polynomial kernel function, Sigmoid core letters
Number and gaussian radial basis function；
Further, dynamic adjustment SVM parametric techniques are taken in the S1, first SVM parameter is determined with priori,
To optimize second SVM parameter, optimizing first SVM parameter after second SVM parameter is determined, finally further according to different ginsengs
Each self-validation of evaluation index result its optimality under several；
Further, the S2 is specially that input sample is tried to achieve Lagrange and multiplied after selected kernel function and SVM parameters
Sub- αi(i=1,2 ..., l) with biasing b, so that it is determined that anticipation function
Further, the S3 is specially to calculate correlation predictive evaluation index, determines predicated error, is normally receiving scope
It is interior, then the road conditions information of prediction is corresponded into corresponding jam level and carry out output displaying；
Further, the S4 is specially：If vi(t) average speed is passed through for the traffic value of t on the i of section
Degree, vi(t-1) it is the traffic value at t-1 moment on the i of section；By xi=[vi(t), vi(t-1) ..., vi(t-m)]TIt is used as sample
The input value of this t, vi(t+1) as the output valve y of sampleI,After the completion of SVM training, by anticipation function formula to traffic road
Condition value is predicted；
Beneficial effects of the present invention are as follows：
1) the nonlinear regression prediction based on SVMs has strict theory and Fundamentals of Mathematics, based on structure risk most
Smallization principle, generalization ability is better than neutral net etc., and algorithm has Global Optimality, according to some training samples, finds one
Expected risk between optimal function, several estimation Y' and reality output Y to inputting X is minimized；
2) method for employing dynamically-adjusting parameter goes to obtain SVM models, kernel function optimized parameter is determined, so as to reach
Precision of prediction is improved, reduces the effect of error；
3) nonlinear regression based on SVM is predicted road conditions value, goes to train SVM according to the road conditions value before adjacent to the period
Model, and then predict the road conditions value of next period, and jam level is converted into according to road conditions value completes each road traffic in short-term and gather around
The prediction of stifled situation.
Brief description of the drawings
Fig. 1 predicts the outcome and real roads road conditions value comparison diagram for heretofore described checking collection.
Embodiment
In order to make the purpose , technical scheme and advantage of the present invention be clearer, it is right below in conjunction with drawings and Examples
The present invention is explained in further detail.It should be appreciated that specific embodiment described herein is used only for explaining the present invention, and
It is not used in the restriction present invention.On the contrary, the present invention cover it is any be defined by the claims the present invention spirit and scope on do
Replacement, modification, equivalent method and scheme.Further, in order that the public has a better understanding to the present invention, below to this
It is detailed to describe some specific detail sections in the detailed description of invention.It is thin without these for a person skilled in the art
The description of section part can also understand the present invention completely.
The invention will be further described with specific embodiment below in conjunction with the accompanying drawings, but not as a limitation of the invention.
Below most preferred embodiment is enumerated for the present invention：
The present invention provides a kind of method predicted based on SVM algorithm traffic in short-term, is specifically based on SVM pre- measuring and calculating
Method step is as follows：
Step1, selects kernel function, determines SVM parameters.After data set is obtained, gaussian kernel function is selected as kernel function,
Optimal Parameters ε, C comprising width parameter δ, quadratic programming.Due to based on priori selection parameter, different pieces of information pair can be caused
Study first appropriate is different, the final overall precision of prediction of influence.Therefore dynamically-adjusting parameter method is taken, it is true with priori
Fixed first parameter, to optimize second parameter, is optimizing first parameter, finally further according to not after second parameter determination
Each self-validation of evaluation index result its optimality under same parameter.
Step2, input data set generates anticipation function.After selected kernel function and SVM parameters, input sample is tried to achieve
Lagrange multipliers αi(i=1,2 ..., l) with biasing b, so that it is determined that anticipation function
Step3, output result after evaluation analysis is carried out according to predicting the outcome.Correlation predictive evaluation index is calculated, it is determined that in advance
Survey error.In the range of normal receive, then the road conditions information of prediction is corresponded into corresponding jam level and carry out output exhibition
Show.
The process of traffic prediction in short-term：
Because traffic flow is the current road conditions and the road conditions of preceding several periods in short-term on consecutive variations, section over time
There is close relation, go the road conditions of prediction future time period using the traffic information of several periods in short-term before section herein accordingly.
If vi(t) average speed, v are passed through for the traffic value of t on the i of sectioni(t-1) when being t-1 on the i of section
The traffic value at quarter.By xi=[vi(t), vi(t-1) ..., vi(t-m)]TIt is used as the input value of sample t, vi(t+1) make
For the output valve y of samplei.After the completion of SVM training, traffic value can be predicted by anticipation function formula.
The SVM algorithm principle of nonlinear regression is as follows：It is assumed that training dataset is designated as T={ (xi,yi)}l i=1, use first
Nonlinear Mapping ψ (x)=[ψ1(x),ψ2(x),…,ψN(x)]TInput data from former space reflection to N-dimensional feature space,
Approximating function is constructed in feature space
The nonlinear regression of lower dimensional space so is correspond in the linear regression of higher dimensional space, definition ε is insensitive loss
Function；
Lε(x, y, f)=| y-f (x) |ε=max (0, | y-f (x) |-ε),
It is exactly to minimize object function to need the nonlinear regression problem solved
Wherein, xi∈RnIt is i-th of input, yi∈ R are corresponding desired outputs；ω=[ω1, ω 2 ..., ω n]TIt is line
Property weights variable, b for biasing；ε and C are two free parameters for having user to determine.
The corresponding outputs of test sample x are predicted as the following formula,
Wherein, K is the kernel function for meeting Mercer conditions, and usually used kernel function has Polynomial kernel function, Sigmoid
Kernel function, gaussian radial basis function.
The present invention is predicted to the road conditions of some wherein one major trunk roads in four-way intersection of a city, and measurement data is
The data of wherein one week, were a measurement interval by every 15 minutes, were continuously counted within daily 24 hours, totally 12542 preparation data.
4 moment road conditions values are used as mode input variable x using current time road conditions value and beforei, wherein in above-mentioned 12542 data
70% as model training sample set, 30% as forecast model checking sample set.It has selected Gauss type kernel functionsGenerally, ε or σ is reduced2Training precision can be improved, but Generalization Capability can be reduced；Increase C
Also training precision can be improved.
Emulation discovery, parameter ε, C, and core width cs2To the considerable influence that predicts the outcome, it is necessary to be entered according to training data
Row is specific to be determined.Figure below is σ2=0.5, C=10, ε=0.05 road conditions value predict the outcome.
Predict the outcome evaluation, and root-mean-square error RMSE is 7.04, and mean absolute error MAE values are 5.39, average absolute hundred
It is 14% to divide error MAPE.
One kind of embodiment described above, simply more preferably embodiment of the invention, those skilled in the art
The usual variations and alternatives that member is carried out in the range of technical solution of the present invention all should be comprising within the scope of the present invention.
Claims (7)
1. a kind of method predicted based on SVM algorithm traffic in short-term, it is characterised in that methods described passes through the non-of SVM
Linear regression is predicted road conditions value, is gone to train SVM models according to the road conditions value of neighbouring period, and then predict the road of next period
Condition value, and the prediction that jam level completes each road traffic congestion situation in short-term is converted into according to road conditions value.
2. according to the method described in claim 1, it is characterised in that methods described includes：
S1：Kernel function is selected, SVM parameters are determined, after the data set of road conditions value of neighbouring period is obtained, gaussian kernel function is selected
As kernel function, the SVM parameters include width parameter δ, the Optimal Parameters ε and C of quadratic programming；
S2:The road conditions Value Data collection of input neighbouring period generates anticipation function as sample；
S3:Output result after evaluation analysis is carried out according to predicting the outcome；
S4:Traffic information according to representated by output result, carries out traffic in short-term and predicts.
3. method according to claim 2, it is characterised in that select kernel function in the S1, determines the specific of SVM parameters
Method is as follows：
1) assume that training dataset is designated as T={ (xi,yi)}l I=1, first with Nonlinear Mapping ψ (x)=[ψ1(x),ψ2(x),…,
ψN(x)]TInput data from former space reflection to N-dimensional feature space, approximating function is constructed in feature space, wherein N is whole
Number；
2) nonlinear regression of lower dimensional space is correspond in the linear regression of higher dimensional space, definition ε is insensitive loss function Lε
(x, y, f)=| y-f (x) |ε=max (0, | y-f (x) |-ε)；
3) object function is minimized
<mrow>
<mfrac>
<mn>1</mn>
<mn>2</mn>
</mfrac>
<mo>|</mo>
<mo>|</mo>
<mi>&omega;</mi>
<mo>|</mo>
<msup>
<mo>|</mo>
<mn>2</mn>
</msup>
<mo>+</mo>
<mi>C</mi>
<munderover>
<mi>&Sigma;</mi>
<mrow>
<mi>i</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>l</mi>
</munderover>
<msub>
<mi>L</mi>
<mi>&epsiv;</mi>
</msub>
<mrow>
<mo>(</mo>
<msub>
<mi>x</mi>
<mi>i</mi>
</msub>
<mo>,</mo>
<msub>
<mi>y</mi>
<mi>i</mi>
</msub>
<mo>,</mo>
<mi>f</mi>
<mo>)</mo>
</mrow>
</mrow>
Wherein, xi∈RnIt is i-th of input, yi∈ R are corresponding desired outputs；ω=[ω1, ω2..., ωn]TIt is linear weight value
Variable, b is biasing；ε and C are two free parameters determined by user；
4) test sample x is predicted as the following formula, and correspondence is exported,
<mrow>
<mi>f</mi>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<munderover>
<mi>&Sigma;</mi>
<mrow>
<mi>i</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>l</mi>
</munderover>
<msub>
<mi>&alpha;</mi>
<mi>i</mi>
</msub>
<mi>K</mi>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>,</mo>
<msub>
<mi>x</mi>
<mi>i</mi>
</msub>
<mo>)</mo>
</mrow>
<mo>+</mo>
<mi>b</mi>
</mrow>
Wherein, K is the kernel function for meeting Mercer conditions, and the kernel function includes Polynomial kernel function, Sigmoid kernel functions and
Gaussian radial basis function.
4. method according to claim 3, it is characterised in that dynamic adjustment SVM parametric techniques are taken in the S1, with elder generation
Test knowledge and determine first SVM parameter, to optimize second SVM parameter, in optimization first after second SVM parameter is determined
SVM parameters, finally further according to each self-validation of evaluation index result its optimality under different parameters.
5. method according to claim 2, it is characterised in that the S2 be specially after selected kernel function and SVM parameters,
Input sample, tries to achieve Lagrange multipliers αi(i=1,2 ..., l) with biasing b, so that it is determined that anticipation function
6. method according to claim 2, it is characterised in that the S3 is specially to calculate correlation predictive evaluation index, really
Determine predicated error, in the range of normal receive, then the road conditions information of prediction is corresponded into corresponding jam level progress defeated
Go out displaying.
7. method according to claim 2, it is characterised in that the S4 is specially：If vi(t) it is t on the i of section
Traffic value passes through average speed, vi(t-1) it is the traffic value at t-1 moment on the i of section；By xi=[vi(t), vi
(t-1) ..., vi (t-m)]TIt is used as the input value of sample t, vi(t+1) as the output valve y of samplei, SVM training completions
Afterwards, traffic value is predicted by anticipation function formula.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710433951.1A CN107170234A (en) | 2017-06-09 | 2017-06-09 | A kind of method predicted based on SVM algorithm traffic in short-term |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710433951.1A CN107170234A (en) | 2017-06-09 | 2017-06-09 | A kind of method predicted based on SVM algorithm traffic in short-term |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107170234A true CN107170234A (en) | 2017-09-15 |
Family
ID=59825158
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710433951.1A Pending CN107170234A (en) | 2017-06-09 | 2017-06-09 | A kind of method predicted based on SVM algorithm traffic in short-term |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107170234A (en) |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107967803A (en) * | 2017-11-17 | 2018-04-27 | 东南大学 | Traffic congestion Forecasting Methodology based on multi-source data and variable-weight combined forecasting model |
| CN108109382A (en) * | 2018-02-05 | 2018-06-01 | 青岛大学 | A kind of congestion points based on composite network, congestion line, the discovery method of congestion regions |
| CN108986453A (en) * | 2018-06-15 | 2018-12-11 | 华南师范大学 | A kind of traffic movement prediction method based on contextual information, system and device |
| CN109754605A (en) * | 2019-02-27 | 2019-05-14 | 中南大学 | A kind of traffic forecast method based on attention temporal diagram convolutional network |
| CN109859477A (en) * | 2019-03-15 | 2019-06-07 | 同盾控股有限公司 | A kind of determination method and apparatus of congestion data |
| CN111915878A (en) * | 2019-05-09 | 2020-11-10 | 奥迪股份公司 | Method and device for predicting road traffic state, computer device and storage medium |
| CN112581753A (en) * | 2019-12-30 | 2021-03-30 | 西安金路交通工程科技发展有限责任公司 | Regional road network dynamic traffic distribution method and system based on omnibearing three-dimensional detection |
| CN112906984A (en) * | 2021-03-24 | 2021-06-04 | 苏州蓝图智慧城市科技有限公司 | Road traffic state prediction method and device, storage medium and electronic equipment |
| CN113096380A (en) * | 2021-03-03 | 2021-07-09 | 南京理工大学 | Short-term road traffic jam prediction method based on BA-SVR algorithm |
| CN113096381A (en) * | 2021-03-03 | 2021-07-09 | 南京理工大学 | Short-term road traffic jam prediction method based on GSA-SVR algorithm |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101541030A (en) * | 2009-05-06 | 2009-09-23 | 华为技术有限公司 | Method for predicting data based on support vector machine and equipment thereof |
| CN103150609A (en) * | 2013-02-18 | 2013-06-12 | 健雄职业技术学院 | Modeling method for short time traffic flow predicting model |
| CN104091216A (en) * | 2014-07-29 | 2014-10-08 | 吉林大学 | Traffic information predication method based on fruit fly optimization least-squares support vector machine |
| US20160314686A1 (en) * | 2013-12-30 | 2016-10-27 | Fudan University | Method for traffic flow prediction based on spatio-temporal correlation mining |
| CN106781465A (en) * | 2016-12-06 | 2017-05-31 | 广州市科恩电脑有限公司 | A kind of road traffic Forecasting Methodology |
-
2017
- 2017-06-09 CN CN201710433951.1A patent/CN107170234A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101541030A (en) * | 2009-05-06 | 2009-09-23 | 华为技术有限公司 | Method for predicting data based on support vector machine and equipment thereof |
| CN103150609A (en) * | 2013-02-18 | 2013-06-12 | 健雄职业技术学院 | Modeling method for short time traffic flow predicting model |
| US20160314686A1 (en) * | 2013-12-30 | 2016-10-27 | Fudan University | Method for traffic flow prediction based on spatio-temporal correlation mining |
| CN104091216A (en) * | 2014-07-29 | 2014-10-08 | 吉林大学 | Traffic information predication method based on fruit fly optimization least-squares support vector machine |
| CN106781465A (en) * | 2016-12-06 | 2017-05-31 | 广州市科恩电脑有限公司 | A kind of road traffic Forecasting Methodology |
Cited By (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107967803A (en) * | 2017-11-17 | 2018-04-27 | 东南大学 | Traffic congestion Forecasting Methodology based on multi-source data and variable-weight combined forecasting model |
| CN108109382A (en) * | 2018-02-05 | 2018-06-01 | 青岛大学 | A kind of congestion points based on composite network, congestion line, the discovery method of congestion regions |
| CN108109382B (en) * | 2018-02-05 | 2020-08-04 | 青岛大学 | Congestion point, congestion line and congestion area discovery method based on composite network |
| CN108986453A (en) * | 2018-06-15 | 2018-12-11 | 华南师范大学 | A kind of traffic movement prediction method based on contextual information, system and device |
| CN109754605B (en) * | 2019-02-27 | 2021-12-07 | 中南大学 | Traffic prediction method based on attention temporal graph convolution network |
| CN109754605A (en) * | 2019-02-27 | 2019-05-14 | 中南大学 | A kind of traffic forecast method based on attention temporal diagram convolutional network |
| CN109859477A (en) * | 2019-03-15 | 2019-06-07 | 同盾控股有限公司 | A kind of determination method and apparatus of congestion data |
| CN109859477B (en) * | 2019-03-15 | 2021-03-19 | 同盾控股有限公司 | Congestion data determination method and device |
| CN111915878A (en) * | 2019-05-09 | 2020-11-10 | 奥迪股份公司 | Method and device for predicting road traffic state, computer device and storage medium |
| CN112581753A (en) * | 2019-12-30 | 2021-03-30 | 西安金路交通工程科技发展有限责任公司 | Regional road network dynamic traffic distribution method and system based on omnibearing three-dimensional detection |
| CN113096380A (en) * | 2021-03-03 | 2021-07-09 | 南京理工大学 | Short-term road traffic jam prediction method based on BA-SVR algorithm |
| CN113096381A (en) * | 2021-03-03 | 2021-07-09 | 南京理工大学 | Short-term road traffic jam prediction method based on GSA-SVR algorithm |
| CN112906984A (en) * | 2021-03-24 | 2021-06-04 | 苏州蓝图智慧城市科技有限公司 | Road traffic state prediction method and device, storage medium and electronic equipment |
| CN112906984B (en) * | 2021-03-24 | 2023-06-30 | 苏州蓝图智慧城市科技有限公司 | Road traffic state prediction method and device, storage medium and electronic equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107170234A (en) | A kind of method predicted based on SVM algorithm traffic in short-term | |
| He et al. | Mining transition rules of cellular automata for simulating urban expansion by using the deep learning techniques | |
| Shao et al. | Traffic flow prediction with long short-term memory networks (LSTMs) | |
| Yu et al. | k-Nearest neighbor model for multiple-time-step prediction of short-term traffic condition | |
| Grekousis et al. | Modeling urban evolution using neural networks, fuzzy logic and GIS: The case of the Athens metropolitan area | |
| Jun | A comparison of a gradient boosting decision tree, random forests, and artificial neural networks to model urban land use changes: the case of the Seoul metropolitan area | |
| Vlahogianni et al. | A real-time parking prediction system for smart cities | |
| Wu et al. | Evolving RBF neural networks for rainfall prediction using hybrid particle swarm optimization and genetic algorithm | |
| CN102693633B (en) | Short-term traffic flow weighted combination prediction method | |
| CN111080029B (en) | Urban traffic road speed prediction method and system based on multi-path segment space-time correlation | |
| CN112541302B (en) | Air quality prediction model training method, air quality prediction method and device | |
| CN109902880A (en) | A kind of city stream of people's prediction technique generating confrontation network based on Seq2Seq | |
| CN109359166B (en) | Space growth dynamic simulation and driving force factor contribution degree synchronous calculation method | |
| Li et al. | Urban water demand forecasting based on HP filter and fuzzy neural network | |
| CN106781489A (en) | A kind of road network trend prediction method based on recurrent neural network | |
| CN106910199A (en) | Towards the car networking mass-rent method of city space information gathering | |
| CN106096756A (en) | A kind of urban road network dynamic realtime Multiple Intersections routing resource | |
| Kişi | Evolutionary fuzzy models for river suspended sediment concentration estimation | |
| Bienvenido-Huertas et al. | Comparison of artificial intelligence algorithms to estimate sustainability indicators | |
| Lai et al. | Integration of multicriteria evaluation and cellular automata methods for landslide simulation modelling | |
| Zhao et al. | Incorporating spatio-temporal smoothness for air quality inference | |
| Naseri et al. | A newly developed hybrid method on pavement maintenance and rehabilitation optimization applying Whale Optimization Algorithm and random forest regression | |
| Jiang et al. | Advanced traffic congestion early warning system based on traffic flow forecasting and extenics evaluation | |
| Li et al. | A method of rainfall runoff forecasting based on deep convolution neural networks | |
| CN107481523A (en) | A kind of traffic flow speed Forecasting Methodology and system |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170915 |