CN109410581A - Traffic flow forecasting method based on wavelet neural network - Google Patents
Traffic flow forecasting method based on wavelet neural network Download PDFInfo
- Publication number
- CN109410581A CN109410581A CN201811372292.6A CN201811372292A CN109410581A CN 109410581 A CN109410581 A CN 109410581A CN 201811372292 A CN201811372292 A CN 201811372292A CN 109410581 A CN109410581 A CN 109410581A
- Authority
- CN
- China
- Prior art keywords
- section
- vehicle
- training
- value
- neural network
- 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
- 238000013528 artificial neural network Methods 0.000 title claims abstract description 15
- 238000013277 forecasting method Methods 0.000 title claims abstract description 11
- 238000012549 training Methods 0.000 claims abstract description 27
- 238000000034 method Methods 0.000 claims abstract description 13
- 230000006870 function Effects 0.000 claims description 25
- 238000004364 calculation method Methods 0.000 claims description 12
- 238000005457 optimization Methods 0.000 claims description 9
- 238000012545 processing Methods 0.000 claims description 6
- 230000000052 comparative effect Effects 0.000 claims description 3
- 238000002513 implantation Methods 0.000 claims description 3
- 238000005259 measurement Methods 0.000 claims description 3
- 238000012552 review Methods 0.000 claims description 3
- 230000000694 effects Effects 0.000 abstract description 6
- 238000010801 machine learning Methods 0.000 abstract description 3
- 230000010355 oscillation Effects 0.000 abstract description 3
- 238000004458 analytical method Methods 0.000 description 14
- 230000009466 transformation Effects 0.000 description 7
- 230000007547 defect Effects 0.000 description 2
- 230000007812 deficiency Effects 0.000 description 2
- 230000004807 localization Effects 0.000 description 2
- 241001269238 Data Species 0.000 description 1
- 230000002159 abnormal effect Effects 0.000 description 1
- 230000004913 activation Effects 0.000 description 1
- 230000002411 adverse Effects 0.000 description 1
- 238000010219 correlation analysis Methods 0.000 description 1
- 235000013399 edible fruits Nutrition 0.000 description 1
- 238000005194 fractionation Methods 0.000 description 1
- 230000001939 inductive effect Effects 0.000 description 1
- 210000004218 nerve net Anatomy 0.000 description 1
- 230000001537 neural effect Effects 0.000 description 1
- 238000003062 neural network model Methods 0.000 description 1
- 210000002569 neuron Anatomy 0.000 description 1
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
- G08G1/0137—Measuring and analyzing of parameters relative to traffic conditions for specific applications
- G08G1/0145—Measuring and analyzing of parameters relative to traffic conditions for specific applications for active traffic flow control
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- 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
- 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
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/40—Business processes related to the transportation industry
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Economics (AREA)
- Human Resources & Organizations (AREA)
- Strategic Management (AREA)
- Health & Medical Sciences (AREA)
- General Business, Economics & Management (AREA)
- General Health & Medical Sciences (AREA)
- Tourism & Hospitality (AREA)
- Marketing (AREA)
- Computational Linguistics (AREA)
- Quality & Reliability (AREA)
- Game Theory and Decision Science (AREA)
- Development Economics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Operations Research (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Entrepreneurship & Innovation (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Primary Health Care (AREA)
- Traffic Control Systems (AREA)
Abstract
The present invention be directed to existing machine learning can fall into oscillation effect due to the sudden situation of road traffic during learning prediction, training process is caused to be difficult to collect and collect increased periods in other words, cause the training time too long, the undesirable situation of training result devises a kind of traffic flow forecasting method based on wavelet neural network thus.
Description
Technical field
The present invention relates to the methods that forecasting traffic flow improves traffic, and in particular to the traffic flow based on wavelet neural network is pre-
Survey method；
Background technique
Wavelet analysis (wavelet Analysis) be a mathematical theory that the 1980s, mid-term grew up and
Method, the proposition when carrying out seismic signal analysis by French scientist Grossman and Morlet then rapidly develop.
Meyer in 1985 demonstrates the existence of wavelet function under one-dimensional situation, and theoretically furthers investigate. Mallat
Based on multiresolution analysis thought, the Mallat algorithm to play an important role to wavelet application is proposed, its ground in wavelet analysis
Status of the position phase group FFT in classical Fourier analysis.The importance of Wavelet Analysis Theory and the popularity of application cause
The great attentions of scientific and technological circle.The appearance of wavelet analysis is considered as the breakthrough of Fourier analysis, in Approximation Theory, micro-
Divide equation, mould to know identification, computer vision, image procossing, nonlinear science etc. to obtain using wavelet analysis in many prominent
Broken property progress.The basic thought of wavelet transformation is converted similar to Fourier, is exactly the sky with signal in cluster basic function
Between on projection characterize the signal.Classical Fourier transformation signal by triangle just, cosine basis be unfolded, by arbitrary function table
Be shown as the linear superposition with the harmonic function of different frequency, can preferably delineate the frequency characteristic of signal, but it in time domain or
Without any resolution on airspace, partial analysis cannot be made.This all brings many deficiencies in theory and using upper.In order to overcome this
One defect, proposes the windowed Fourier transform.By inducing one, a time localization " window function " improves Fourier transformation
Deficiency, but its window size and shape are fixed, and do not make up the defect of Fourier transformation fundamentally.And small echo
Transformation has good localization performance simultaneously in time domain and frequency domain, has the T/F window an of flexibility and changeability, this is resonable
By and practical application it is all significant.Wavelet neural is the neural network model constituted based on wavelet transformation with network, that is, is used
Nonlinear wavelet base replaces common neuron nonlinear activation function (such as Sigmoid function), wavelet transformation and nerve net
The characteristics of network organically combines, and sufficiently inherits the two.
Summary of the invention
The present invention be directed to existing machine learning during learning prediction due to the sudden situation of road traffic
Oscillation effect is fallen into, causes training process to be difficult to collect and collect increased periods in other words, causes the training time too long, training knot
The undesirable situation of fruit devises a kind of traffic flow forecasting method based on wavelet neural network thus.
Traffic flow forecasting method based on wavelet neural network, comprising the following steps:
M1 sets multichannel CCD identification, using FGPA as the prediction meanss of single channel processing core；
M2 predicts 5 traffic lights jump periods or 15 minutes, the correct then jump procedure of prediction so that device carries out study prediction；
M3 establishes training set and is written data are acquired in step M1；
M4 carries out five groups of training to data, is divided into 1 traffic lights jump period or 3 minutes between training；
M5, statistical forecast output collects degree, judges whether emergency case occur, without emergency case then jump procedure M7；
M6, transfers history hands-on data in the same time, and whether review emergency case is true；
M7, comparative training prediction result and actual result, error rate is less than setting value, and then training of judgement is completed, and training is terminated
Algorithm model launch application.
Preferably, the step M4 the following steps are included:
A1 reads leading to section and flowing into section or take relevance higher 3 for current road segment on the basis of current road segment mouth
Flow into section；
A2 transfers the wagon flow magnitude of the different periods of road section selected in A1；
A3, the prediction vehicle flowrate Q after a timing section τ is calculated according to formula
Formula: Q (t+ τ)=F ({ Q (t-k1 τ) }, { Qd (t-k2 τ) }, { Qu (t-k3 τ) })
Wherein: Q is vehicle flowrate counting, and F is to calculate function, and t is the selected period, and τ is timing section, QuTo flow into section vehicle flowrate
It counts, QdIt is counted to lead to section vehicle flowrate, k1, k2, k3 is the measurement degree of correlation and is integer；
A4 carries out the processing of Mexico's hat wavelet function to input, formula according to the input of step A3:
A5 obtains contraction-expansion factor aij and shift factor bij；
A6 obtains result using minimum operationCalculation formula:
A7 is calculatedWhereinIt is network connection weight.
Preferably, the step M4 the following steps are included:
B1 calculates the vehicle flowrate Q (t-k1 τ) run on road section selected r to intersection c；
B2, calculates the practical bout length l of section r, and calculating vehicle calculates on the r of section to intersection c speed of service V (t)
To runing time t of the vehicle on the section；
B3, according to the ratio calculation intersection delay value di of green time and red time；
B4, according to point of runing time t and intersection delay value di calculating vehicle on the r of section in step B2 and step B3
Implantation rn；
B5 calculates the vehicle flowrate output valve Qd (t- k2 τ) of the vehicle flowrate input value Qu (t-k3 τ) and r-1 of section r+1 according to rn；
B6 passes through formula calculating target function
Wherein QmnIt is m that (t-k1 τ), which is state, when n, the wagon flow magnitude of section r, and QdmnIt is m that (t-k2 τ), which is state, when n, section r's
Vehicle flowrate output valve, QumnIt is m that (t-k3 τ), which is state, when n, the vehicle flowrate input value of section r；
B7 repeats step B1 to B6 until step B6 statistical value and actual value difference ratio are less than setting value.
Preferably, the step B6 the following steps are included:
Function is inherited in B61, optimizationWherein f is to calculate target, and N is to calculate race's size；
B62, brings the functional value in B6 into, and calculation optimization inherits collection of functions；
B63 selects optimization to inherit in collection of functions numerical value maximum 3 and as optimized calculation method brings latter wheel into.
Preferably, should also select the vehicle flowrate data under emergency case, sample choosing in addition to conventional statistic numerical value is brought into
It is selected as the current vehicle distribution and the distribution of the vehicle at latter two moment of section r, input sample should include the pass come out
First 3 in the highest section of connection degree.
Substantial effect of the invention is adding due to wavelet neural network during the adjustment in weight and threshold value
Momentum term can be effectively reduced the oscillation effect of learning process, improve convergence rate, improve training effect；Preferably tracking is handed over
Through-flow data characteristics reduces the adverse effect of abnormal data, improves precision of prediction.
Specific embodiment
Below by specific embodiment, technical scheme of the present invention will be further explained in detail.
Embodiment 1
The traffic flow forecasting method based on wavelet neural network, comprising the following steps:
M1 sets multichannel CCD identification, using FGPA as the prediction meanss of single channel processing core；
M2 predicts 5 traffic lights jump periods or 15 minutes, the correct then jump procedure of prediction so that device carries out study prediction；
M3 establishes training set and is written data are acquired in step M1；
M4 carries out five groups of training to data, is divided into 1 traffic lights jump period or 3 minutes between training；
M5, statistical forecast output collects degree, judges whether emergency case occur, without emergency case then jump procedure M7；
M6, transfers history hands-on data in the same time, and whether review emergency case is true；
M7, comparative training prediction result and actual result, error rate is less than setting value, and then training of judgement is completed, and training is terminated
Algorithm model launch application.
In forecasting traffic flow, the traffic flow data of several continuous times before relevant road segments cannot be simply picked to carry out
Prediction needs to carry out the traffic datas of different sections of highway, different periods correlation analysis to improve precision of prediction, and these
The magnitude of traffic flow have time-varying and nonlinear characteristic, be difficult to provide more accurate analytical expression, it is therefore desirable to traffic flow data into
Row confluence analysis generally carries out automatic training analysis by the way of machine learning, and the present invention has then selected and used small wavelength-division
The neural network method of analysis；Wherein at least it is divided into three layers by the fractionation to model, input layer is that multiple groups training data rule is
3 classes are respectively current road segment and are generally the traffic flow of current road segment with other biggish two sections of current road segment relevance
Input section and output section；Computation layer is then related to that small echo calculates or neural network transmitting calculates, and basic includes blurring, mould
Paste matching is subordinate to calculating and deblurring；Output layer is then related to minimum operation, for matching primitives layer calculated result and output
The matching state of layer, and obtain actual computation model.
Preferably, the step M4 the following steps are included:
A1 reads leading to section and flowing into section or take relevance higher 3 for current road segment on the basis of current road segment mouth
Flow into section；
A2 transfers the wagon flow magnitude of the different periods of road section selected in A1；
A3, the prediction vehicle flowrate Q after a timing section τ is calculated according to formula
Formula: Q (t+ τ)=F ({ Q (t-k1 τ) }, { Qd (t-k2 τ) }, { Qu (t-k3 τ) })
Wherein: Q is vehicle flowrate counting, and F is to calculate function, and t is the selected period, and τ is timing section, QuTo flow into section vehicle flowrate
It counts, QdIt is counted to lead to section vehicle flowrate, k1, k2, k3 is the measurement degree of correlation and is integer；
A4 carries out the processing of Mexico's hat wavelet function to input, formula according to the input of step A3:
A5 obtains contraction-expansion factor aij and shift factor bij；
A6 obtains result using minimum operationCalculation formula:
A7 is calculatedWhereinIt is network connection weight.
The optimal speed that this step is improved in practice needs rule to calculate between two o'clock all in city
Operating path and runing time are to reduce the intermediate waiting time.
Preferably, the step M4 the following steps are included:
B1 calculates the vehicle flowrate Q (t-k1 τ) run on road section selected r to intersection c；
B2, calculates the practical bout length l of section r, and calculating vehicle calculates on the r of section to intersection c speed of service V (t)
To runing time t of the vehicle on the section；
B3, according to the ratio calculation intersection delay value di of green time and red time；
B4, according to point of runing time t and intersection delay value di calculating vehicle on the r of section in step B2 and step B3
Implantation rn；
B5 calculates the vehicle flowrate output valve Qd (t- k2 τ) of the vehicle flowrate input value Qu (t-k3 τ) and r-1 of section r+1 according to rn；
B6 passes through formula calculating target function
Wherein QmnIt is m that (t-k1 τ), which is state, when n, the wagon flow magnitude of section r, and QdmnIt is m that (t-k2 τ), which is state, when n, section r's
Vehicle flowrate output valve, QumnIt is m that (t-k3 τ), which is state, when n, the vehicle flowrate input value of section r；
B7 repeats step B1 to B6 until step B6 statistical value and actual value difference ratio are less than setting value.
Preferably, the step B6 the following steps are included:
Function is inherited in B61, optimizationWherein f is to calculate target, and N is to calculate race's size；
B62, brings the functional value in B6 into, and calculation optimization inherits collection of functions；
B63 selects optimization to inherit in collection of functions numerical value maximum 3 and as optimized calculation method brings latter wheel into.
Preferably, should also select the vehicle flowrate data under emergency case, sample choosing in addition to conventional statistic numerical value is brought into
It is selected as the current vehicle distribution and the distribution of the vehicle at latter two moment of section r, input sample should include the pass come out
First 3 in the highest section of connection degree.
Claims (5)
1. the traffic flow forecasting method based on wavelet neural network, which comprises the following steps:
M1 sets multichannel CCD identification, using FGPA as the prediction meanss of single channel processing core；
M2 predicts 5 traffic lights jump periods or 15 minutes, the correct then jump procedure of prediction so that device carries out study prediction；
M3 establishes training set and is written data are acquired in step M1；
M4 carries out five groups of training to data, is divided into 1 traffic lights jump period or 3 minutes between training；
M5, statistical forecast output collects degree, judges whether emergency case occur, without emergency case then jump procedure M7；
M6, transfers history hands-on data in the same time, and whether review emergency case is true；
M7, comparative training prediction result and actual result, error rate is less than setting value, and then training of judgement is completed, and training is terminated
Algorithm model launch application.
2. the traffic flow forecasting method according to claim 1 based on wavelet neural network, which is characterized in that the step
Rapid M4 the following steps are included:
A1 reads leading to section and flowing into section or take relevance higher 3 for current road segment on the basis of current road segment mouth
Flow into section；
A2 transfers the wagon flow magnitude of the different periods of road section selected in A1；
A3, the prediction vehicle flowrate Q after a timing section τ is calculated according to formula
Formula: Q (t+ τ)=F ({ Q (t-k1 τ) }, { Qd (t-k2 τ) }, { Qu (t-k3 τ) })
Wherein: Q is vehicle flowrate counting, and F is to calculate function, and t is the selected period, and τ is timing section, QuTo flow into section wagon flow meter
Number, QdIt is counted to lead to section vehicle flowrate, k1, k2, k3 is the measurement degree of correlation and is integer；
A4 carries out the processing of Mexico's hat wavelet function to input, formula according to the input of step A3:
A5 obtains contraction-expansion factor aij and shift factor bij；
A6 obtains result using minimum operationCalculation formula:
A7 is calculatedWhereinIt is network connection weight.
3. the traffic flow forecasting method according to claim 1 based on wavelet neural network, which is characterized in that the step
Rapid M4 the following steps are included:
B1 calculates the vehicle flowrate Q (t-k1 τ) run on road section selected r to intersection c；
B2, calculates the practical bout length l of section r, and calculating vehicle calculates on the r of section to intersection c speed of service V (t)
To runing time t of the vehicle on the section；
B3, according to the ratio calculation intersection delay value di of green time and red time；
B4, according to point of runing time t and intersection delay value di calculating vehicle on the r of section in step B2 and step B3
Implantation rn；
B5 calculates the vehicle flowrate output valve Qd (t-k2 τ) of the vehicle flowrate input value Qu (t-k3 τ) and r-1 of section r+1 according to rn；
B6 passes through formula calculating target function
Wherein QmnIt is m that (t-k1 τ), which is state, when n, the wagon flow magnitude of section r, and QdmnIt is m that (t-k2 τ), which is state, when n, the vehicle of section r
Flow output valve, QumnIt is m that (t-k3 τ), which is state, when n, the vehicle flowrate input value of section r；
B7 repeats step B1 to B6 until step B6 statistical value and actual value difference ratio are less than setting value.
4. the traffic flow forecasting method according to claim 3 based on wavelet neural network, which is characterized in that the step
Rapid B6 the following steps are included:
Function is inherited in B61, optimizationWherein f is to calculate target, and N is to calculate race's size；
B62, brings the functional value in B6 into, and calculation optimization inherits collection of functions；
B63 selects optimization to inherit in collection of functions numerical value maximum 3 and as optimized calculation method brings latter wheel into.
5. the traffic flow forecasting method according to claim 2 based on wavelet neural network, which is characterized in that except conventional system
Count value is brought into outer, should also select the vehicle flowrate data under emergency case, the current vehicle distribution that samples selection is section r with
And the vehicle distribution at latter two moment, input sample should include first 3 in the highest section of the degree of association come out.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811372292.6A CN109410581A (en) | 2018-11-16 | 2018-11-16 | Traffic flow forecasting method based on wavelet neural network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811372292.6A CN109410581A (en) | 2018-11-16 | 2018-11-16 | Traffic flow forecasting method based on wavelet neural network |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109410581A true CN109410581A (en) | 2019-03-01 |
Family
ID=65473994
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811372292.6A Pending CN109410581A (en) | 2018-11-16 | 2018-11-16 | Traffic flow forecasting method based on wavelet neural network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109410581A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110009429A (en) * | 2019-04-10 | 2019-07-12 | 金瓜子科技发展（北京）有限公司 | A kind of method, apparatus and computer equipment of predicted flow rate data |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CA2513928A1 (en) * | 2003-01-22 | 2004-08-05 | Matsushita Electric Industrial Co., Ltd. | A traffic information providing system, a traffic information representation method and apparatus therefor |
| CN101127158A (en) * | 2006-08-18 | 2008-02-20 | 株式会社查纳位资讯情报 | Predictive traffic information creating method, predictive traffic information creating apparatus, and traffic information display terminal |
| CN103914981A (en) * | 2014-04-09 | 2014-07-09 | 江苏物联网研究发展中心 | Method for predicting confliction between pedestrians and left-turn vehicles at plane intersection |
| CN108346293A (en) * | 2018-04-18 | 2018-07-31 | 湖南大学 | A kind of arithmetic for real-time traffic flow Forecasting Approach for Short-term |
-
2018
- 2018-11-16 CN CN201811372292.6A patent/CN109410581A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CA2513928A1 (en) * | 2003-01-22 | 2004-08-05 | Matsushita Electric Industrial Co., Ltd. | A traffic information providing system, a traffic information representation method and apparatus therefor |
| CN101127158A (en) * | 2006-08-18 | 2008-02-20 | 株式会社查纳位资讯情报 | Predictive traffic information creating method, predictive traffic information creating apparatus, and traffic information display terminal |
| CN103914981A (en) * | 2014-04-09 | 2014-07-09 | 江苏物联网研究发展中心 | Method for predicting confliction between pedestrians and left-turn vehicles at plane intersection |
| CN108346293A (en) * | 2018-04-18 | 2018-07-31 | 湖南大学 | A kind of arithmetic for real-time traffic flow Forecasting Approach for Short-term |
Non-Patent Citations (1)
| Title |
|---|
| 邵俊倩: "基于小波模糊神经网络的实时交通流预测", 《计算机***应用》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110009429A (en) * | 2019-04-10 | 2019-07-12 | 金瓜子科技发展（北京）有限公司 | A kind of method, apparatus and computer equipment of predicted flow rate data |
| CN110009429B (en) * | 2019-04-10 | 2021-04-16 | 金瓜子科技发展（北京）有限公司 | Method and device for predicting flow data and computer equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111210633B (en) | Short-term traffic flow prediction method based on deep learning | |
| CN106846538B (en) | Cross vehicle record treating method and apparatus | |
| CN101789182B (en) | Traffic signal control system and method based on parallel simulation technique | |
| CN109064748B (en) | Traffic average speed prediction method based on time cluster analysis and variable convolutional neural network | |
| CN104077613B (en) | Crowd density estimation method based on cascaded multilevel convolution neural network | |
| CN109214581B (en) | Railway line wind speed prediction method considering wind direction and confidence interval | |
| CN105321345B (en) | A kind of road traffic flow prediction method filtered based on ARIMA models and kalman | |
| CN106384509A (en) | Urban road driving time distribution estimation method considering taxi operation states | |
| CN107832882A (en) | A kind of taxi based on markov decision process seeks objective policy recommendation method | |
| CN114023062B (en) | Traffic flow information monitoring method based on deep learning and edge calculation | |
| CN112071062B (en) | Driving time estimation method based on graph convolution network and graph attention network | |
| CN111929748A (en) | Meteorological element forecasting method and system | |
| CN112907970B (en) | Variable lane steering control method based on vehicle queuing length change rate | |
| CN107705556A (en) | A kind of traffic flow forecasting method combined based on SVMs and BP neural network | |
| CN110047291A (en) | A kind of Short-time Traffic Flow Forecasting Methods considering diffusion process | |
| CN105913658B (en) | A kind of method that traffic flow speculates OD positions and OD matrixes | |
| EP3726434B1 (en) | Method for reducing exhaust emissions of a vehicle drive system with combustion engine | |
| CN106504548A (en) | Traffic lights intelligent control method and system | |
| CN109191845A (en) | A kind of public transit vehicle arrival time prediction technique | |
| CN109118787A (en) | A kind of car speed prediction technique based on deep neural network | |
| CN108648445A (en) | Dynamic traffic Tendency Prediction method based on traffic big data | |
| CN105551250A (en) | Method for discriminating urban road intersection operation state on the basis of interval clustering | |
| CN113449905A (en) | Traffic jam early warning method based on gated cyclic unit neural network | |
| CN111524345B (en) | Induction control method for multi-objective optimization under constraint of real-time queuing length of vehicle | |
| CN111341109A (en) | City-level signal recommendation system based on space-time similarity |
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: 20190301 |