CN110070715A - A kind of road traffic flow prediction method based on Conv1D-NLSTMs neural network structure - Google Patents

A kind of road traffic flow prediction method based on Conv1D-NLSTMs neural network structure Download PDF

Info

Publication number
CN110070715A
CN110070715A CN201910352267.XA CN201910352267A CN110070715A CN 110070715 A CN110070715 A CN 110070715A CN 201910352267 A CN201910352267 A CN 201910352267A CN 110070715 A CN110070715 A CN 110070715A
Authority
CN
China
Prior art keywords
traffic flow
road
road traffic
data
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
Application number
CN201910352267.XA
Other languages
Chinese (zh)
Inventor
徐东伟
朱钟华
彭鹏
王永东
戴宏伟
魏臣臣
宣琦
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Zhejiang University of Technology ZJUT
Original Assignee
Zhejiang University of Technology ZJUT
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Zhejiang University of Technology ZJUT filed Critical Zhejiang University of Technology ZJUT
Priority to CN201910352267.XA priority Critical patent/CN110070715A/en
Publication of CN110070715A publication Critical patent/CN110070715A/en
Pending legal-status Critical Current

Links

Classifications

    • 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
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/045—Combinations of networks
    • 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
    • G06N3/084—Backpropagation, e.g. using gradient descent
    • 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
    • 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

Landscapes

  • Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • General Health & Medical Sciences (AREA)
  • General Engineering & Computer Science (AREA)
  • Biophysics (AREA)
  • Computational Linguistics (AREA)
  • Data Mining & Analysis (AREA)
  • Evolutionary Computation (AREA)
  • Artificial Intelligence (AREA)
  • Molecular Biology (AREA)
  • Computing Systems (AREA)
  • Biomedical Technology (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Health & Medical Sciences (AREA)
  • Chemical & Material Sciences (AREA)
  • Analytical Chemistry (AREA)
  • Traffic Control Systems (AREA)

Abstract

A kind of road traffic flow prediction method based on Conv1D-NLSTMs neural network structure, comprising the following steps: 1), construct the traffic flow data matrix of associated road, and data are pre-processed；2) road traffic flow space-time characteristic, is extracted based on traffic flow data matrix；3) road traffic flow prediction model, is constructed based on road traffic flow space-time characteristic: regression forecasting is made to obtained road traffic flow space-time characteristic using full articulamentum, future time instance is obtained without the prediction result of the road traffic flow of renormalization, and model parameter is continued to optimize using back-propagation algorithm according to the result of mean square error, result is mapped as actual traffic flow value eventually by renormalization；4) it, verifies road traffic flow prediction model: the road traffic flow data in test set being predicted using the model that training is completed, compare prediction result and actual value to test model performance.Prediction result of the present invention is more acurrate according to the experimental results.

Description

A kind of road traffic flow prediction based on Conv1D-NLSTMs neural network structure Method
Technical field
The invention belongs to traffic forecast fields, are related to a kind of road friendship based on Conv1D-NLSTMs neural network structure Through-flow prediction technique.
Background technique
With the high speed development of social economy, domestic car ownership is significantly increased, but the following traffic is stifled Plug problem also brings great inconvenience to the trip of people.And the appearance of intelligent transportation system, various traffic datas are carried out Analysis carries out management regulation real-time, accurately and efficiently to traffic above-ground to realize, alleviates ground friendship to a certain extent Logical pressure.And road traffic flow predicts a part as intelligent transportation system, anticipated that the traffic in following a period of time State plays important function to the real-time management for realizing traffic.
Existing road traffic flow prediction method is mainly based upon the prediction model or right of mathematical statistics, machine learning Several class models carry out single combination, although certain prediction effect can be obtained, there are still some limitations, often neglect Certain characteristics slightly in traffic flow data.
Summary of the invention
In order to overcome the lower deficiency of precision of existing road traffic forecast method, the present invention provides one kind and is based on The road traffic flow prediction method of Conv1D-NLSTMs neural network structure, the method use the traffic of a plurality of associated road Flow data is therefrom extracted the space characteristics of road traffic flow, meanwhile, pervious traffic flow data also can be to future the long period The magnitude of traffic flow at moment generates certain influence, and the NLSTMs neural network in this method can enhance the feature extraction of this respect.
The technical solution adopted by the present invention to solve the technical problems is:
A kind of road traffic flow prediction method based on Conv1D-NLSTMs neural network structure, comprising the following steps:
1) the traffic flow data matrix of associated road, is constructed, and data are pre-processed: selection prediction road and its phase The traffic flow data of associated road constructs traffic flow data matrix, and it is normalized；
2) road traffic flow space-time characteristic, is extracted based on traffic flow data matrix: using one-dimensional convolutional network to traffic flow The traffic flow data of synchronization different sections of highway extracts space characteristics in data matrix, obtains the sequence number with space characteristics According to reusing NLSTMs neural network and extract temporal aspect in the sequence data, to obtain road traffic flow space-time characteristic；
3) road traffic flow prediction model, is constructed based on road traffic flow space-time characteristic: using full articulamentum to obtaining Road traffic flow space-time characteristic makees regression forecasting, obtains future time instance without the prediction knot of the road traffic flow of renormalization Fruit, and model parameter is continued to optimize using back-propagation algorithm according to the result of mean square error, it will be tied eventually by renormalization Fruit is mapped as actual traffic flow value；
4), verify road traffic flow prediction model: the model completed using training is to the road traffic flow in test set According to being predicted, prediction result and actual value are compared to test model performance.
Technical concept of the invention are as follows: mainly utilize one-dimensional convolutional network (Conv1D) and Nested LSTMs (NLSTMs) Neural network, extracts feature in terms of the spatial relationship of road traffic flow and the time series of traffic data two, and Conv1D is mentioned The space characteristics of road traffic flow data have been taken, and NLSTMs neural network is similar to and is internally embedded in LSTM neural network structure One or more LSTM neural networks, can extraction time sequence signature more more effective than LSTM neural network, to improve The precision of prediction of road traffic flow
Beneficial effects of the present invention: one-dimensional convolutional neural networks can effectively be mentioned from the traffic flow data of a plurality of associated road Space characteristics are taken, and the reinforcement of NLSTMs neural network considers longer time pervious traffic flow to future time instance traffic flow It influences, the variation tendency of road traffic flow is excavated from multiple angles, so that prediction result is more acurrate.
Detailed description of the invention
Fig. 1 is NLSTMs neural network structure figure；
Fig. 2 is Conv1D-NLSTMs Artificial Neural Network Structures figure；
Fig. 3 is the forecasting traffic flow result based on Conv1D-NLSTMs neural network model.
Specific embodiment
The invention will be further described below in conjunction with the accompanying drawings.
Referring to Fig.1~Fig. 3, a kind of road traffic flow prediction method based on Conv1D-NLSTMs neural network structure, packet Include following steps:
1) the traffic flow data matrix of associated road, is constructed, and data are pre-processed, process is as follows:
First, it is assumed that predicting the magnitude of traffic flow of road k, then the traffic flow of the road and its associated road is chosen Data are measured, for constructing traffic flow data matrix X ', form is as follows:
Wherein, column vector respectively represents the traffic flow sequence data of m different sections of highway, and row vector represents different road n The traffic flow data at moment, then x 'itFor the original traffic flow data of t moment road i；
According to the time series of road traffic flow, training data and test data, division proportion 8:2 are splitted data into.
Then road traffic flow data is normalized using maxmin criterion method, process is as follows:
Wherein, x 'max, x 'minMaximin respectively in original traffic flow data, xitFor t moment road after pretreatment The magnitude of traffic flow of road i, enables Pt=[x1t,x2t,…,xmt], indicate the traffic flow modes of t moment difference road, then by pretreatment Traffic flow data matrix X afterwards are as follows:
2) road traffic flow space-time characteristic is extracted based on traffic flow data matrix；
To the P at each momenttMake one-dimensional convolution operation, extracts the space characteristics of road traffic flow data, calculating process It is as follows:
st=f (Wp*Pt+bp) (4)
Wherein, WpIndicate weight matrix, bpIndicate bias term, * indicates that convolution algorithm, f indicate activation primitive relu:max { x, 0 }, stIndicate that one-dimensional convolution algorithm obtains as a result, the space characteristics of the road traffic flow data then extracted are S=[s1, s2,…,sn]T；
Before space characteristics are input to NLSTMs neural network, need to make space characteristics S formal transformation, Variation is as follows:
Wherein, d indicates to take d continuous data before road traffic flow to predict the magnitude of traffic flow of subsequent time, S '=[xd, xd+1,…,xn-1]T, n >=d, it is after representation space eigentransformation as a result, then the input of NLSTMs neural network be xt,d≤t≤n- 1, and its sample number is n-d；
NLSTMs neural network is divided into inside and outside 2 parts, the update of external LSTM neural network location mode and gate Mechanism is expressed as following equation:
it=σ (xtWxi+ht-1Whi+bi) (6)
ft=σ (xtWxf+ht-1Whf+bf) (7)
ot=σ (xtWxo+ht-1Who+bo) (9)
ht=ot·σ(ct) (10)
Wherein, dot product is indicated, σ () indicates sigmoid function, Wxf、Wxi、WxoIt indicates external and forgets door, input gate, defeated The input weight matrix gone out, Whf、Whi、WhoIndicate the external previous moment output weight square for forgeing door, input gate, out gate Battle array, bf、bi、boIndicate the external bias matrix for forgeing door, input gate, out gate, it、ft、ct、ot、htExpression external input door, Forget the output of door, location mode, out gate, memory unit,Then indicate the output of memory internal unit；
The state of its calculation formula for being internally embedded LSTM neural network unit and LSTM neural network updates and gate machine Calculation formula processed is similar, and expression formula is as follows:
Wherein, dot product is indicated, σ () indicates sigmoid function,Indicate internal input, WxcIndicate internal input Weight matrix, WhcIndicate the internal previous moment state cell weight matrix inputted, bcIndicate the internal bias matrix inputted,Indicate the internal input weight matrix for forgeing door, input gate, state cell, out gate, Indicate the internal previous moment output weight for forgeing door, input gate, state cell, out gate Matrix,Indicate the internal bias matrix for forgeing door, input gate, state cell, out gate,The output for indicating internal input gate, forgeing door, location mode, out gate, memory unit.Therefore The final output of NLSTMs neural network, i.e. road traffic flow space-time characteristic are H=ht；
3) road traffic flow prediction model is constructed based on road traffic flow space-time characteristic, process is as follows:
Firstly, inputting road traffic flow space-time characteristic H as full articulamentum, prediction is based under input traffic flow data One moment magnitude of traffic flow yt(without renormalization), full connection expression formula are as follows:
yt=Wh·H (18)
Wherein, WhFor the weight matrix of full articulamentum, and ytCorresponding true value Yt=xK, t+1,d≤t≤n-1；
Then defining mean square error is loss function L:
Then the loss function L of computation model continues to optimize model parameter using back-propagation algorithm realization, reversely Gradient in propagation algorithm is calculated to be realized by Adam optimizer with parameter update；
Finally, by the output y of full articulamentumtMake renormalization operation, actual traffic flow forecasting value can be obtained；
4) road traffic flow prediction model is verified
Model is verified using test data, prediction result and actual value are made comparisons.Absolute value is chosen in this experiment The index of mean square deviation (MAE), root-mean-square error (RMSE) as road traffic flow precision of prediction, calculation formula distinguish following institute Show:
Wherein, b is sample number, YoFor actual traffic flow, Yo' the predicted flow rate exported for model.
Example: the data in actual experiment, prediction process are as follows:
1) experimental data is chosen
Original road traffic flow data includes 3 roads, 29 days traffic flow datas, which is north Capital city bicyclic moieties link flow data, sampling interval T are 2min.23 days before this 3 roads road traffic flow datas are made For training data, model parameter training is carried out, rear 6 days road traffic flow datas carry out model performance and test as test data Card.
2) parameter determines
Experiment of the invention is realized based on tensorflow environment, completes entire experimental model frame using keras Frame is built, and one-dimensional convolution process is realized by the Conv1D function in keras, and NLSTMs neural network passes through NestedLSTM Layer realizes that full articulamentum is realized by Dense function.Therefore entire experiment parameter set it is as follows: the number of plies of one-dimensional convolution is defeated as 1 Entering matrix size is that (road sum is 3 to 10x3, and the data on flows at former 10 moment is predicted that both d=10, convolution kernel were long Degree 1, convolution nuclear volume 3, filling mode is " padding " and activation primitive is relu:max { x, 0 }；NestedLSTM layers defeated Unit is 64 out, and the number of plies is set as 2；Full articulamentum output unit quantity is 1, that is, predicts the magnitude of traffic flow of subsequent time.
3) experimental result
In an experiment, the magnitude of traffic flow of this 3 roads is predicted respectively, while by this method and LSTM nerve net Network, NLSTMs neural network method compare, and as a result statistical analysis is as shown in table 1:
Table 1.

Claims (5)

1. a kind of road traffic flow prediction method based on Conv1D-NLSTMs neural network structure, which is characterized in that the side Method the following steps are included:
(1), the traffic flow data matrix of associated road is constructed, and data are pre-processed: selection prediction road and its correlation Join the traffic flow data of road, constructs traffic flow data matrix, and it is normalized；
(2), road traffic flow space-time characteristic is extracted based on traffic flow data matrix: using one-dimensional convolutional network to traffic flow data The traffic flow data of synchronization different sections of highway extracts space characteristics in matrix, obtains the sequence data with space characteristics, then The temporal aspect in the sequence data is extracted using NLSTMs neural network, to obtain road traffic flow space-time characteristic；
(3), road traffic flow prediction model is constructed based on road traffic flow space-time characteristic: using full articulamentum to obtained road Traffic flow space-time characteristic makees regression forecasting, obtains future time instance without the prediction result of the road traffic flow of renormalization, and Model parameter is continued to optimize using back-propagation algorithm according to the result of mean square error, maps result eventually by renormalization For actual traffic flow value；
(4), verify road traffic flow prediction model: using training complete model to the road traffic flow data in test set into Row prediction compares prediction result and actual value to test model performance.
2. a kind of road traffic flow prediction method based on Conv1D-NLSTMs neural network structure as described in claim 1, It is characterized in that, the process of the step (1) is as follows:
First, it is assumed that predicting the magnitude of traffic flow of road k, then the magnitude of traffic flow number of the road and its associated road is chosen According to for constructing traffic flow data matrix X ', form is as follows:
Wherein, column vector respectively represents the traffic flow sequence data of m different sections of highway, and row vector represents the n moment of different roads Traffic flow data, then x 'itFor the original traffic flow data of t moment road i；
According to the time series of road traffic flow, training data and test data, division proportion 8:2 are splitted data into；
Then road traffic flow data is normalized using maxmin criterion method, process is as follows:
Wherein, x 'max, x 'minMaximin respectively in original traffic flow data, xitFor t moment road i after pretreatment The magnitude of traffic flow enables Pt=[x1t,x2t,…,xmt], indicate the traffic flow modes of t moment difference road, then by pretreated Traffic flow data matrix X are as follows:
3. a kind of road traffic flow prediction side based on Conv1D-NLSTMs neural network structure as claimed in claim 1 or 2 Method, which is characterized in that in the step (2), to the P at each momenttMake one-dimensional convolution operation, extracts road traffic flow data Space characteristics, calculating process are as follows:
st=f (Wp*Pt+bp) (4)
Wherein, WpIndicate weight matrix, bpIndicating bias term, * indicates that convolution algorithm, f indicate activation primitive relu:max { x, 0 }, stIndicate that one-dimensional convolution algorithm obtains as a result, the space characteristics of the road traffic flow data then extracted are S=[s1,s2,…, sn]T；
It before space characteristics are input to NLSTMs neural network, needs to make space characteristics S formal transformation, converts Form is as follows:
Wherein, d indicates to take d continuous data before road traffic flow to predict the magnitude of traffic flow of subsequent time, S '=[xd, xd+1,…,xn-1]T, n >=d, it is after representation space eigentransformation as a result, then the input of NLSTMs neural network be xt,d≤t≤n- 1, and its sample number is n-d；
NLSTMs neural network is divided into inside and outside 2 parts, the update of external LSTM neural network location mode and door control mechanism It is expressed as following equation:
it=σ (xtWxi+ht-1Whi+bi) (6)
ft=σ (xtWxf+ht-1Whf+bf) (7)
ot=σ (xtWxo+ht-1Who+bo) (9)
ht=ot·σ(ct) (10)
Wherein, dot product is indicated, σ () indicates sigmoid function, Wxf、Wxi、WxoIt indicates external and forgets door, input gate, out gate Input weight matrix, Whf、Whi、WhoIndicate the external previous moment output weight matrix for forgeing door, input gate, out gate, bf、bi、boIndicate the external bias matrix for forgeing door, input gate, out gate, it、ft、ct、ot、htIt indicates external input door, forget The output of door, location mode, out gate, memory unit,Then indicate the output of memory internal unit；
The state of its calculation formula for being internally embedded LSTM neural network unit and LSTM neural network updates and door control mechanism meter Calculation formula is similar, and expression formula is as follows:
Wherein, dot product is indicated, σ () indicates sigmoid function,Indicate internal input, WxcIndicate the internal weight inputted Matrix, WhcIndicate the internal previous moment state cell weight matrix inputted, bcIndicate the internal bias matrix inputted,Indicate the internal input weight matrix for forgeing door, input gate, state cell, out gate, Indicate the internal previous moment output weight for forgeing door, input gate, state cell, out gate Matrix,Indicate the internal bias matrix for forgeing door, input gate, state cell, out gate,The output for indicating internal input gate, forgeing door, location mode, out gate, memory unit, therefore The final output of NLSTMs neural network, i.e. road traffic flow space-time characteristic are H=ht。
4. a kind of road traffic flow prediction side based on Conv1D-NLSTMs neural network structure as claimed in claim 1 or 2 Method, which is characterized in that the process of the step (3) is as follows:
Firstly, inputting road traffic flow space-time characteristic H as full articulamentum, lower a period of time based on input traffic flow data is predicted Carve magnitude of traffic flow yt, it is as follows that expression formula is connected entirely:
yt=Wh·H (18)
Wherein, WhFor the weight matrix of full articulamentum, and ytCorresponding true value Yt=xK, t+1,d≤t≤n-1；
Then defining mean square error is loss function L:
Then the loss function L of computation model continues to optimize model parameter using back-propagation algorithm realization, backpropagation Gradient in algorithm is calculated to be realized by Adam optimizer with parameter update；
Finally, by the output y of full articulamentumtMake renormalization operation, actual traffic flow forecasting value can be obtained.
5. a kind of road traffic flow prediction side based on Conv1D-NLSTMs neural network structure as claimed in claim 1 or 2 Method, which is characterized in that in the step (4), choose absolute value mean square deviation MAE, root-mean-square error RMSE as road traffic flow The index of precision of prediction, calculation formula difference are as follows:
Wherein, b is sample number, YoFor actual traffic flow, Yo' the predicted flow rate exported for model.
CN201910352267.XA 2019-04-29 2019-04-29 A kind of road traffic flow prediction method based on Conv1D-NLSTMs neural network structure Pending CN110070715A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201910352267.XA CN110070715A (en) 2019-04-29 2019-04-29 A kind of road traffic flow prediction method based on Conv1D-NLSTMs neural network structure

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201910352267.XA CN110070715A (en) 2019-04-29 2019-04-29 A kind of road traffic flow prediction method based on Conv1D-NLSTMs neural network structure

Publications (1)

Publication Number Publication Date
CN110070715A true CN110070715A (en) 2019-07-30

Family

ID=67369433

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201910352267.XA Pending CN110070715A (en) 2019-04-29 2019-04-29 A kind of road traffic flow prediction method based on Conv1D-NLSTMs neural network structure

Country Status (1)

Country Link
CN (1) CN110070715A (en)

Cited By (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111009129A (en) * 2020-01-08 2020-04-14 武汉大学 Urban road traffic flow prediction method and device based on space-time deep learning model
CN111292525A (en) * 2019-12-30 2020-06-16 南京艾特斯科技有限公司 Traffic flow prediction method based on neural network
CN111639787A (en) * 2020-04-28 2020-09-08 北京工商大学 Spatio-temporal data prediction method based on graph convolution network
CN111862592A (en) * 2020-05-27 2020-10-30 浙江工业大学 Traffic flow prediction method based on RGCN
CN112530157A (en) * 2020-10-16 2021-03-19 浙江工业大学 Road traffic congestion propagation prediction method based on knowledge graph and Conv1D-LSTM-D
CN112906945A (en) * 2021-01-27 2021-06-04 昭通亮风台信息科技有限公司 Traffic flow prediction method, system and computer readable storage medium
CN112990598A (en) * 2021-03-31 2021-06-18 浙江禹贡信息科技有限公司 Reservoir water level time sequence prediction method and system
CN113159372A (en) * 2021-01-29 2021-07-23 太原理工大学 Conv1D + LSTM-based multi-step traffic flow prediction method
CN114120652A (en) * 2021-12-21 2022-03-01 重庆邮电大学 Traffic flow prediction method based on dynamic graph neural network
CN114724386A (en) * 2022-03-31 2022-07-08 温州大学 Short-time traffic flow prediction method and system under intelligent traffic and electronic equipment
CN115116226A (en) * 2022-06-24 2022-09-27 云控智行(上海)汽车科技有限公司 Highway freight car flow prediction algorithm, software and device based on deep reinforcement learning model
CN115691151A (en) * 2022-12-29 2023-02-03 浙江大华技术股份有限公司 Lane congestion prompting method, lane congestion prompting system and terminal
CN116244281A (en) * 2022-09-28 2023-06-09 北京百度网讯科技有限公司 Lane traffic flow data complement and model training method and device thereof

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105389980A (en) * 2015-11-09 2016-03-09 上海交通大学 Short-time traffic flow prediction method based on long-time and short-time memory recurrent neural network
CN108399749A (en) * 2018-03-14 2018-08-14 西南交通大学 A kind of traffic trip needing forecasting method in short-term
CN108510741A (en) * 2018-05-24 2018-09-07 浙江工业大学 A kind of traffic flow forecasting method based on Conv1D-LSTM neural network structures
CN108647834A (en) * 2018-05-24 2018-10-12 浙江工业大学 A kind of traffic flow forecasting method based on convolutional neural networks structure
KR101951595B1 (en) * 2018-05-18 2019-02-22 한양대학교 산학협력단 Vehicle trajectory prediction system and method based on modular recurrent neural network architecture
CN109410575A (en) * 2018-10-29 2019-03-01 北京航空航天大学 A kind of road network trend prediction method based on capsule network and the long Memory Neural Networks in short-term of nested type
CN109544911A (en) * 2018-10-30 2019-03-29 中山大学 A kind of city road net traffic state prediction technique based on LSTM-CNN

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105389980A (en) * 2015-11-09 2016-03-09 上海交通大学 Short-time traffic flow prediction method based on long-time and short-time memory recurrent neural network
CN108399749A (en) * 2018-03-14 2018-08-14 西南交通大学 A kind of traffic trip needing forecasting method in short-term
KR101951595B1 (en) * 2018-05-18 2019-02-22 한양대학교 산학협력단 Vehicle trajectory prediction system and method based on modular recurrent neural network architecture
CN108510741A (en) * 2018-05-24 2018-09-07 浙江工业大学 A kind of traffic flow forecasting method based on Conv1D-LSTM neural network structures
CN108647834A (en) * 2018-05-24 2018-10-12 浙江工业大学 A kind of traffic flow forecasting method based on convolutional neural networks structure
CN109410575A (en) * 2018-10-29 2019-03-01 北京航空航天大学 A kind of road network trend prediction method based on capsule network and the long Memory Neural Networks in short-term of nested type
CN109544911A (en) * 2018-10-30 2019-03-29 中山大学 A kind of city road net traffic state prediction technique based on LSTM-CNN

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
NICHOLAS G. POLSON: "Deep learning for short-term traffic flow prediction", 《TRANSPORTATION RESEARCH PART C: EMERGING TECHNOLOGIES》 *
康丹青: "基于深度学习的短时交通流预测方法研究", 《中国优秀硕士学位论文全文数据库工程科技Ⅱ辑》 *

Cited By (19)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111292525A (en) * 2019-12-30 2020-06-16 南京艾特斯科技有限公司 Traffic flow prediction method based on neural network
CN111292525B (en) * 2019-12-30 2023-01-13 南京艾特斯科技有限公司 Traffic flow prediction method based on neural network
CN111009129B (en) * 2020-01-08 2021-06-15 武汉大学 Urban road traffic flow prediction method and device based on space-time deep learning model
CN111009129A (en) * 2020-01-08 2020-04-14 武汉大学 Urban road traffic flow prediction method and device based on space-time deep learning model
CN111639787A (en) * 2020-04-28 2020-09-08 北京工商大学 Spatio-temporal data prediction method based on graph convolution network
CN111639787B (en) * 2020-04-28 2024-03-15 北京工商大学 Spatio-temporal data prediction method based on graph convolution network
CN111862592A (en) * 2020-05-27 2020-10-30 浙江工业大学 Traffic flow prediction method based on RGCN
CN112530157B (en) * 2020-10-16 2022-05-03 浙江工业大学 Road traffic congestion propagation prediction method based on knowledge graph and Conv1D-LSTM-D
CN112530157A (en) * 2020-10-16 2021-03-19 浙江工业大学 Road traffic congestion propagation prediction method based on knowledge graph and Conv1D-LSTM-D
CN112906945A (en) * 2021-01-27 2021-06-04 昭通亮风台信息科技有限公司 Traffic flow prediction method, system and computer readable storage medium
CN113159372A (en) * 2021-01-29 2021-07-23 太原理工大学 Conv1D + LSTM-based multi-step traffic flow prediction method
CN112990598A (en) * 2021-03-31 2021-06-18 浙江禹贡信息科技有限公司 Reservoir water level time sequence prediction method and system
CN114120652A (en) * 2021-12-21 2022-03-01 重庆邮电大学 Traffic flow prediction method based on dynamic graph neural network
CN114724386A (en) * 2022-03-31 2022-07-08 温州大学 Short-time traffic flow prediction method and system under intelligent traffic and electronic equipment
CN114724386B (en) * 2022-03-31 2023-10-27 温州大学 Short-time traffic flow prediction method and system under intelligent traffic and electronic equipment
CN115116226A (en) * 2022-06-24 2022-09-27 云控智行(上海)汽车科技有限公司 Highway freight car flow prediction algorithm, software and device based on deep reinforcement learning model
CN116244281A (en) * 2022-09-28 2023-06-09 北京百度网讯科技有限公司 Lane traffic flow data complement and model training method and device thereof
CN116244281B (en) * 2022-09-28 2023-11-21 北京百度网讯科技有限公司 Lane traffic flow data complement and model training method and device thereof
CN115691151A (en) * 2022-12-29 2023-02-03 浙江大华技术股份有限公司 Lane congestion prompting method, lane congestion prompting system and terminal

Similar Documents

Publication Publication Date Title
CN110070715A (en) A kind of road traffic flow prediction method based on Conv1D-NLSTMs neural network structure
CN108510741A (en) A kind of traffic flow forecasting method based on Conv1D-LSTM neural network structures
CN110021165A (en) A kind of traffic flow forecasting method based on Autoencoder-LSTM Fusion Model
CN112418547B (en) Bus stop passenger flow prediction method based on GCN-LSTM combination model
CN105654729B (en) A kind of short-term traffic flow forecast method based on convolutional neural networks
CN113053115B (en) Traffic prediction method based on multi-scale graph convolution network model
CN111292525B (en) Traffic flow prediction method based on neural network
Pan et al. AutoSTG: Neural Architecture Search for Predictions of Spatio-Temporal Graph✱
Shao et al. Traffic flow prediction with long short-term memory networks (LSTMs)
CN109492822A (en) Air pollutant concentration time-space domain interaction prediction method
CN103761138B (en) Parameter correction method for traffic simulation software
CN106295799B (en) A kind of implementation method of deep learning multilayer neural network
CN108664687A (en) A kind of industrial control system space-time data prediction technique based on deep learning
CN110223517A (en) Short-term traffic flow forecast method based on temporal correlation
CN108009674A (en) Air PM2.5 concentration prediction methods based on CNN and LSTM fused neural networks
CN107665230A (en) Training method and device for the users' behavior model of Intelligent housing
CN106815971A (en) A kind of landslide disaster forecasting procedure based on RBF neural
CN110516833A (en) A method of the Bi-LSTM based on feature extraction predicts road traffic state
CN102622418A (en) Prediction device and equipment based on BP (Back Propagation) nerve network
CN110164129B (en) Single-intersection multi-lane traffic flow prediction method based on GERNN
CN114780739B (en) Time sequence knowledge graph completion method and system based on time graph convolution network
CN111523706B (en) Section lane-level short-term traffic flow prediction method based on deep learning combination model
CN103927891A (en) Crossroad dynamic turning proportion two-step prediction method based on double Bayes
CN102622515A (en) Weather prediction method
CN107704970A (en) A kind of Demand-side load forecasting method based on Spark

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: 20190730