CN105321356A - 2-dimensional flow prediction based intelligent urban traffic-flow prediction method - Google Patents
2-dimensional flow prediction based intelligent urban traffic-flow prediction method Download PDFInfo
- Publication number
- CN105321356A CN105321356A CN201510901367.5A CN201510901367A CN105321356A CN 105321356 A CN105321356 A CN 105321356A CN 201510901367 A CN201510901367 A CN 201510901367A CN 105321356 A CN105321356 A CN 105321356A
- Authority
- CN
- China
- Prior art keywords
- dimension
- data
- prediction
- horizontal
- traffic flow
- 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
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Traffic Control Systems (AREA)
Abstract
The invention provides a 2-dimensional flow prediction based intelligent urban traffic-flow prediction method, which comprises the following steps: firstly, carrying out horizontal-dimension and vertical-dimension sampling on historical data, so that horizontal-dimension data and vertical-dimension data are obtained; secondly, by using the horizontal-dimension data and the vertical-dimension data, for horizontal-dimension and vertical-dimension, respectively selecting different prediction algorithms to carry out traffic prediction, so that horizontal-dimension prediction data and vertical-dimension prediction data are respectively obtained; and finally, by using a 2-dimensional traffic flow prediction algorithm, merging the horizontal-dimension prediction data and the vertical-dimension prediction data, so that a final flow prediction value is obtained.
Description
Technical field
The present invention relates to urban road arithmetic for real-time traffic flow prediction field, more particularly, the present invention relates to a kind of intelligent city's traffic flow forecasting method based on the prediction of 2-dimension stream.
Background technology
Along with the quickening of urban construction and the progressively development of intelligent transportation system, intelligent traffic administration system and advanced urban transportation forecast system play more and more important effect in urban traffic control.In the recent period, many cities, especially in some big cities, the stream of people or the crowded generation all causing some accidents of traffic flow, therefore, forecasting traffic flow technology seems especially important.
European Union have selected the exemplarily cities, city such as Paris, Rome, Madrid, using the exploitation of traffic forecast instrument as important content wherein in its CAPITALSplus project.Britain develops the function of traffic forecast in its following 10aITS project TrafficEngland, can provide the Prediction Parameters of main roads future time period.The project development website of BayernInfo of the BayernOnline that Bavarian state government of Germany initiates, one of its major function provides the forecasting traffic flow information of long, medium and short phase for traveler exactly, have employed the traffic model of a by name/ASDA2FOTO0.The U.S. has some states, city studying and are setting up travel forecasting systems.Domestic scholars has done certain research to arithmetic for real-time traffic flow Short-term Forecasting Model, but there is no successful story at present.
At present, most prediction algorithm belongs to 1-and ties up prediction algorithm, major defect to produce a step lag issues, namely carry out prediction produce the time point of acute variation under traffic flow bursty state after and can produce larger error, current most of 1 dimension prediction algorithm all also exists such problem, especially in magnitude of traffic flow complexity or when there is acute variation.
Summary of the invention
Technical matters to be solved by this invention is for there is above-mentioned defect in prior art, provides a kind of and can solve the intelligent city's traffic flow forecasting method based on the prediction of 2-dimension stream that 1-ties up the step lag issues occurred in prediction algorithm.
In order to realize above-mentioned technical purpose, according to the present invention, provide a kind of intelligent city's traffic flow forecasting method based on the prediction of 2-dimension stream, comprising: first, historical data is carried out to the sampling of horizontal dimension and vertical dimension both direction with the data of the data and vertical dimension that obtain horizontal dimensions; Secondly, utilize the data of horizontal dimensions and the data of vertical dimension, select different prediction algorithms to carry out volume forecasting respectively for horizontal dimension and vertical dimension, to obtain horizontal dimension predicted data and vertical dimension predicted data respectively; Finally, utilize 2-to tie up forecasting traffic flow algorithm horizontal dimension predicted data and vertical dimension predicted data are merged to obtain final flow rate predicted value.
Preferably, wavelet neural network is adopted to carry out volume forecasting for horizontal dimension.
Preferably, autoregression integration moving average model is adopted to carry out volume forecasting for vertical dimension.
Preferably, horizontal dimensions data according to every day different time change carry out adding up producing.
Preferably, vertical dimension data according to not on the same day the data of same time carry out adding up producing.
The present invention can be effectively applied to urban traffic flow (comprising Vehicle flow or the stream of people etc.) prediction aspect, effectively can reduce crowded or trample the generation of accident.
Accompanying drawing explanation
By reference to the accompanying drawings, and by reference to detailed description below, will more easily there is more complete understanding to the present invention and more easily understand its adjoint advantage and feature, wherein:
Fig. 1 schematically shows according to the preferred embodiment of the invention based on the horizontal dimension data instance of intelligent city's traffic flow forecasting method of 2-dimension stream prediction.
Fig. 2 schematically shows according to the preferred embodiment of the invention based on the vertical dimension data instance of intelligent city's traffic flow forecasting method of 2-dimension stream prediction.
Fig. 3 schematically shows according to the preferred embodiment of the invention based on the process flow diagram of intelligent city's traffic flow forecasting method of 2-dimension stream prediction.
It should be noted that, accompanying drawing is for illustration of the present invention, and unrestricted the present invention.Note, represent that the accompanying drawing of structure may not be draw in proportion.Further, in accompanying drawing, identical or similar element indicates identical or similar label.
Embodiment
In order to make content of the present invention clearly with understandable, below in conjunction with specific embodiments and the drawings, content of the present invention is described in detail.
The feature of urban traffic flow historical data can describe from horizontal dimension and vertical dimension two aspects.The data of horizontal dimensions carry out adding up producing according to the change of different time every day, the data of vertical dimension then according to not on the same day the data of same time carry out adding up producing.
If Δ T and Δ t represents the time interval of diurnal periodicity and data sampling respectively.Here Δ T pro data, is 24 hours, then in d days, the traffic flow data of t time period can be expressed as f (t, d), wherein t ∈ [1, Δ T/ Δ t], and d ∈ [1, h], h is number of days total in historical data base here.Such as f (3,5) represents the sampling of the 3rd time period of in historical data the 5th day.
Horizontal dimension the time period different according to every day in historical data base is carried out statistics obtain.Fig. 1 is that Kodaira dimension is according to distribution example.If x={f ' (d), d=1,2 ..., n} and f ' (d)=and f (t, d) | t=1,2 ..., Δ T/ Δ t}, then Kodaira dimension is according to being expressed as:
X＝{x
i|i＝1,2,...,Lx}(1)
Here Lx=n Δ T/ Δ t (n≤h) length that is horizontal dimension.
The data statistics of not going up at one time on the same day in the data representation historical data base of vertical dimensions.If
Y
t=f (t, d) (d=1,2 ..., Ly, Ly≤h), then the data of vertical dimension can be expressed as
Y＝{y
(t,i)|i＝1,2,...,Ly}(2)
Here Ly is the length of vertical dimension.
Treatment scheme based on intelligent city's traffic flow forecasting method of 2-dimension stream prediction is as follows: first, historical data is carried out to the sampling of horizontal dimension and vertical dimension both direction; Secondly, select different prediction algorithms to carry out volume forecasting respectively for horizontal dimension and vertical dimension; Finally, utilize 2-to tie up forecasting traffic flow algorithm and the predicted data merging of horizontal dimension and vertical dimension is obtained final flow rate predicted value.
Fig. 3 schematically shows according to the preferred embodiment of the invention based on the process flow diagram of intelligent city's traffic flow forecasting method of 2-dimension stream prediction.
< horizontal dimension >
Kodaira dimension is according to the situation of change illustrating each time period flow in a day, and characteristic comprises less violent data on flows change at random, is therefore applicable to adopting nonlinear model to carry out data modeling.The preferred embodiment of the present invention adopts wavelet neural network to carry out Kodaira dimension and it is predicted.Wavelet neural network combines the advantage of neural network and wavelet analysis, is widely used in Forecast of Nonlinear Time Series.The key step of horizontal dimension traffic flow data prediction is as follows:
1. horizontal dimension data on flows normalization, if x
ifor the flow of moment i corresponding in horizontal dimension, then it can be normalized to
Here v
max=1, v
min=-1, x
maxrepresent the maximum magnitude of traffic flow in horizontal dimension, x
minrepresent the minimum magnitude of traffic flow in horizontal dimension,
2. determine the nodes of neural network every layer.Wavelet neural network has input layer, hidden layer and output layer 3 layers.
If N is input layer number, H is node in hidden layer, and M is output layer nodes, then node in hidden layer can calculate according to the following formula:
Because only the next time period flow of prediction, therefore gets M=1.
3. train wavelet neural network.By constantly regulating input layer and hidden layer, the weights of hidden layer and output layer, and the coefficient of wavelet function for training, neural network finally reaches convergence, obtains the neural network model for testing.
4. the neural network model being used for testing is carried out horizontal dimension traffic flow forecasting, at time t, one group of data for testing are { x
i| i=t, t-1 ..., t-N+1}, exports the predicted value x ' that can obtain time t+1
t+1.
< vertical dimension >
Vertical dimension does not flow quantitative statistics at one time on the same day, the sampling period be one day namely 24 hours because the time cycle is longer, therefore may comprise random variation and the fluctuation of mass data, be applicable to adopting linear modeling approach.The preferred embodiment of the present invention adopts autoregression integration moving average model (ARIMA) method to carry out vertical dimension according to model and forecast.Autoregression integration moving average model is famous Time Series Forecasting Methods, its main thought to be passed in time by forecasting object and the data sequence formed is considered as a random series, with mathematical model ARIMA (p, d, q) this sequence of approximate description is carried out, here p is autoregression item, and q is moving average item number, the difference order of d for doing when time series becomes steady.Adopt the key step of autoregression integration moving average model prediction vertical dimension traffic flow as follows:
1.ARIMA (p, d, q) model is stationary time series definition, therefore first needs the stationarity of first testing vertical dimension certificate, if belong to non-stationary model, then needs to do d jump and assign to obtain a stable time series.If vertical dimension data sequence is y
(t, i)(i=1,2 ..., Ly), then the stationary time series after the differential transformation of d rank can be designated as Δ y
(t, i)(i=1,2 ..., Ly-d), Δ y here
tcomputing method are as follows:
Here
coefficient of autocorrelation, θ
i(i=1,2 ..., q) be partial correlation coefficient.
2. determine p and q in ARIMA (p, d, q) model, autoregression item p and moving average item number q can be obtained by the autocorrelogram and partial autocorrelation figure checking sequence (4).
3. parameter estimation obtains coefficient of autocorrelation
with partial correlation coefficient θ
i(i=1,2 ..., q).
4. utilize the data of ARIMA (p, d, q) model to vertical dimension obtained to predict, obtain vertical dimension predicted value y '
t+1.
< merges >
Horizontal dimension and vertical dimension forecasting traffic flow out after, 2-just can be used to tie up forecasting traffic flow algorithm and two data are carried out merging to obtain optimum prediction value.It is as follows that 2-ties up forecasting traffic flow algorithm steps:
Horizontal dimension predicted value x during t computing time
t' and vertical dimension predicted value y
t'; Weight w=0 is set, predicted value p is set
t=0; And perform subsequent treatment for the i in the scope of 1 to 100:
p
t＝p
t+wx
t′+(1-w)y
t′。
w＝w+0.01
p
t＝p
t/100
Finally export final predicted value p
t.
In a word, current most of Traffic volume forecasting algorithm is confined to the prediction of 1 dimension data.The present invention ties up Forecasting Methodology by proposing a kind of 2-, predicting traffic flow prediction is carried out from transverse direction (horizontal dimension) and longitudinal (vertical dimension) time point, again the result of bidimensional prediction is merged and obtain final predicted data, effectively can solve the step lag issues existed in 1 dimension prediction.
The present invention proposes a kind of 2-and ties up forecasting traffic flow algorithm, utilize 2-to tie up forecasting traffic flow algorithm and extract historical data, in transverse direction (horizontal dimension) with longitudinally, (vertical dimension) time point is predicted respectively, then the result of prediction is carried out merging and obtain final predicted data.In the present invention, horizontal dimension and vertical dimension prediction have employed different forecast models respectively according to features, i.e. wavelet-neural network model and autoregression integration moving average model.Vertical dimension forecast model of the present invention can solve a step lag issues effectively.
Be understandable that, although the present invention with preferred embodiment disclose as above, but above-described embodiment and be not used to limit the present invention.For any those of ordinary skill in the art, do not departing under technical solution of the present invention ambit, the technology contents of above-mentioned announcement all can be utilized to make many possible variations and modification to technical solution of the present invention, or be revised as the Equivalent embodiments of equivalent variations.Therefore, every content not departing from technical solution of the present invention, according to technical spirit of the present invention to any simple modification made for any of the above embodiments, equivalent variations and modification, all still belongs in the scope of technical solution of the present invention protection.
Claims (5)
1., based on intelligent city's traffic flow forecasting method of 2-dimension stream prediction, it is characterized in that comprising: first, historical data is carried out to the sampling of horizontal dimension and vertical dimension both direction with the data of the data and vertical dimension that obtain horizontal dimensions; Secondly, utilize the data of horizontal dimensions and the data of vertical dimension, select different prediction algorithms to carry out volume forecasting respectively for horizontal dimension and vertical dimension, to obtain horizontal dimension predicted data and vertical dimension predicted data respectively; Finally, utilize 2-to tie up forecasting traffic flow algorithm horizontal dimension predicted data and vertical dimension predicted data are merged to obtain final flow rate predicted value.
2. the intelligent city's traffic flow forecasting method based on the prediction of 2-dimension stream according to claim 1, is characterized in that, adopt wavelet neural network to carry out volume forecasting for horizontal dimension.
3. the intelligent city's traffic flow forecasting method based on the prediction of 2-dimension stream according to claim 1 and 2, is characterized in that, adopts autoregression integration moving average model to carry out volume forecasting for vertical dimension.
4. the intelligent city's traffic flow forecasting method based on the prediction of 2-dimension stream according to claim 1 and 2, it is characterized in that, the data of horizontal dimensions carry out adding up producing according to the change of different time every day.
5. the intelligent city's traffic flow forecasting method based on the prediction of 2-dimension stream according to claim 1 and 2, is characterized in that, the data of vertical dimension according to not on the same day the data of same time carry out adding up producing.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510901367.5A CN105321356A (en) | 2015-12-08 | 2015-12-08 | 2-dimensional flow prediction based intelligent urban traffic-flow prediction method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510901367.5A CN105321356A (en) | 2015-12-08 | 2015-12-08 | 2-dimensional flow prediction based intelligent urban traffic-flow prediction method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN105321356A true CN105321356A (en) | 2016-02-10 |
Family
ID=55248655
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510901367.5A Pending CN105321356A (en) | 2015-12-08 | 2015-12-08 | 2-dimensional flow prediction based intelligent urban traffic-flow prediction method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105321356A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110517479A (en) * | 2018-05-22 | 2019-11-29 | 杭州海康威视系统技术有限公司 | A kind of urban highway traffic prediction technique, device and electronic equipment |
| CN113538904A (en) * | 2021-06-22 | 2021-10-22 | 浙江工业大学 | Traffic accident identification method based on space-time preamble coding and decoding predictor |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20060106599A1 (en) * | 2004-11-16 | 2006-05-18 | Microsoft Corporation | Precomputation and transmission of time-dependent information for varying or uncertain receipt times |
| US20070208492A1 (en) * | 2006-03-03 | 2007-09-06 | Inrix, Inc. | Dynamic time series prediction of future traffic conditions |
| CN101290713A (en) * | 2008-07-11 | 2008-10-22 | 浙江工业大学 | Fractal traffic flow prediction method combining weekly similarity |
| CN102034350A (en) * | 2009-09-30 | 2011-04-27 | 北京四通智能交通系统集成有限公司 | Short-time prediction method and system of traffic flow data |
-
2015
- 2015-12-08 CN CN201510901367.5A patent/CN105321356A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20060106599A1 (en) * | 2004-11-16 | 2006-05-18 | Microsoft Corporation | Precomputation and transmission of time-dependent information for varying or uncertain receipt times |
| US20070208492A1 (en) * | 2006-03-03 | 2007-09-06 | Inrix, Inc. | Dynamic time series prediction of future traffic conditions |
| CN101290713A (en) * | 2008-07-11 | 2008-10-22 | 浙江工业大学 | Fractal traffic flow prediction method combining weekly similarity |
| CN102034350A (en) * | 2009-09-30 | 2011-04-27 | 北京四通智能交通系统集成有限公司 | Short-time prediction method and system of traffic flow data |
Non-Patent Citations (3)
| Title |
|---|
| 余国强: "《基于小波神经网络的短时交通流预测算法的研究》", 《中国优秀硕士学位论文全文数据库工程科技Ⅱ辑》 * |
| 关占荣: "《城市高速公路交通流预测与控制》", 《中国优秀硕士学位论文全文数据库工程科技Ⅱ辑》 * |
| 袁健，等: "《城市道路短期交通流预测VHSSA模型》", 《公路交通科技》 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110517479A (en) * | 2018-05-22 | 2019-11-29 | 杭州海康威视系统技术有限公司 | A kind of urban highway traffic prediction technique, device and electronic equipment |
| CN110517479B (en) * | 2018-05-22 | 2020-11-03 | 杭州海康威视系统技术有限公司 | Urban road traffic prediction method and device and electronic equipment |
| CN113538904A (en) * | 2021-06-22 | 2021-10-22 | 浙江工业大学 | Traffic accident identification method based on space-time preamble coding and decoding predictor |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Shu et al. | A short-term traffic flow prediction model based on an improved gate recurrent unit neural network | |
| Hou et al. | An adaptive hybrid model for short-term urban traffic flow prediction | |
| CN105513337B (en) | The Forecasting Methodology and device of a kind of volume of the flow of passengers | |
| Mingheng et al. | Accurate multisteps traffic flow prediction based on SVM | |
| CN105046956B (en) | Traffic flow simulating and predicting method based on turning probability | |
| Yu et al. | A special event-based K-nearest neighbor model for short-term traffic state prediction | |
| CN111815046B (en) | Traffic flow prediction method based on deep learning | |
| CN105513350A (en) | Time-phased multi-parameter short-term traffic flow prediction method based on time-space characteristics | |
| CN104217250A (en) | Rail transit new line opening passenger flow prediction method based on historical data | |
| Yang et al. | Application of data science technologies in intelligent prediction of traffic congestion | |
| Zhi-Peng et al. | An improved adaptive exponential smoothing model for short-term travel time forecasting of urban arterial street | |
| Duan et al. | Forecasting the Short‐Term Traffic Flow in the Intelligent Transportation System Based on an Inertia Nonhomogenous Discrete Gray Model | |
| Hamad et al. | Near‐term travel speed prediction utilizing Hilbert–Huang transform | |
| CN109816983A (en) | A kind of short-term traffic flow forecast method based on depth residual error network | |
| Zhang et al. | Hybrid deep spatio-temporal models for traffic flow prediction on holidays and under adverse weather | |
| CN105321356A (en) | 2-dimensional flow prediction based intelligent urban traffic-flow prediction method | |
| Lu et al. | Optimizing signal timing of the arterial‐branch intersection: A fuzzy control and nonlinear programming approach | |
| Xue et al. | Expressway traffic flow prediction model based on Bi-LSTM neural networks | |
| CN113408786A (en) | Traffic characteristic prediction method and system applicable to long-time prediction based on control information | |
| Wang et al. | Dynamic traffic congestion simulation and dissipation control based on traffic flow theory model and neural network data calibration algorithm | |
| CN106203731A (en) | A kind of stock trend forecasting method based on 2 dimension stream predictions | |
| Gao et al. | Method of Predicting Passenger Flow in Scenic Areas Considering Multisource Traffic Data. | |
| Qin et al. | Traffic flow prediction based on JANet and attention mechanism | |
| Jakteerangkool et al. | Short-term travel time prediction from GPS trace data using recurrent neural networks | |
| Agarwal et al. | Person-based dynamic traffic assignment for mixed traffic conditions |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WD01 | Invention patent application deemed withdrawn after publication | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20160210 |