CN110428613A - A kind of intelligent transportation trend prediction method of machine learning - Google Patents
A kind of intelligent transportation trend prediction method of machine learning Download PDFInfo
- Publication number
- CN110428613A CN110428613A CN201910614671.XA CN201910614671A CN110428613A CN 110428613 A CN110428613 A CN 110428613A CN 201910614671 A CN201910614671 A CN 201910614671A CN 110428613 A CN110428613 A CN 110428613A
- Authority
- CN
- China
- Prior art keywords
- meteorological
- traffic
- model
- attribute
- data collection
- 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
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- 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
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Software Systems (AREA)
- Human Resources & Organizations (AREA)
- Analytical Chemistry (AREA)
- Chemical & Material Sciences (AREA)
- Strategic Management (AREA)
- Economics (AREA)
- Artificial Intelligence (AREA)
- Medical Informatics (AREA)
- General Business, Economics & Management (AREA)
- Game Theory and Decision Science (AREA)
- Quality & Reliability (AREA)
- Operations Research (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Tourism & Hospitality (AREA)
- Development Economics (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Marketing (AREA)
- Entrepreneurship & Innovation (AREA)
- Traffic Control Systems (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a kind of intelligent transportation trend prediction methods of machine learning.The present invention is according to road net data collection, the magnitude of traffic flow-speed data collection, meteorological dataset and the social property data set being collected into, the Encoder-Decoder mixed model based on LSTM is established, the intelligence prediction in real time of traffic behavior is realized by the training to the mixed model.The present invention has fully considered time and space effect, and increases meteorological attribute, road network attribute, social property, is able to ascend the accuracy of road traffic prediction of speed；Wherein in the generating process of meteorological attribute, traffic speed and meteorological degree of correlation in the cell region divided are calculated on the basis of road junction, a possibility that traffic is not by meteorological effect in certain unit areas is considered, therefore the prediction accuracy to all unit areas can be improved.
Description
Technical field
The present invention relates to data minings and machine learning field, and in particular to a kind of intelligent transportation state of machine learning is pre-
Survey method.
Background technique
Intelligent transportation system (ITS) is a kind of extensive advanced technology, it is intended to by the effect for improving traffic operation and management
Rate improves the sustainability of existing transportation network.An importance of ITS is advanced traveler information systems, in order to make to hand over
Logical management system effectively works, and is vital in relation to practical and immediate future traffic behavior information.This is just needed pair
The ability that following traffic flow forecasting is constantly updated.
Traffic forecast has been the important component of intelligent transportation system (ITS) at present, to Transportation Network Planning, routing
Navigate and avoid many applications such as congestion most important.For the big cities such as Beijing, such prediction is most important, but holds
Row gets up challenging.This is because the limited potential of metropolitan dynamic and complicated traffic environment and new road: this
It has just laid stress in the management of traffic.This management has wide influence, not only due to what is be related to is populous, and
Also as it supports the decision of various other applications, such as optimization pollution, urban planning and the relevant application of energy consumption.Cause
This, traffic forecast has considerable effect for city intelligent process.
In real life, influence of the weather conditions for traffic behavior is very big.For example, due to low visibility, vehicle
Demand is big, and heavy rain can slow down traffic speed, cause traffic jam；In very arctic weather, the decline of temperature can make
Icy road influences transport performance；Rainfall influences traffic condition, and then influences the volume of traffic etc. of major urban arterial highway.Therefore, in order to
Improve the traffic administration under different weather situation, the intelligent transportation trend prediction method for developing a kind of consideration weather information is very
Significant.
At present in machine learning field, maximum hot spot is unquestionably deep learning, on processing large data collection, tool
There are extraordinary learning ability and prediction effect.Wherein, LSTM model achieves thrilling in terms of sequence data prediction
Effect, therefore just it is being suitably applied the prediction of intelligent transportation state.LSTM model is earliest by Hochreiter and Schmidhuber
It was proposed in 1997, is for solving a kind of certain variations that gradient disappears in RNN model or gradient expands.By in RNN
Multiple thresholdings are introduced, so that different moments integral can change, so that gradient be avoided to disappear in the case where model parameter is fixed
Or expansion issues.
An existing technology is " a kind of aerodrome traffic congestion prediction technique and device based on LSTM model ".For prediction
The congestion index of road in the preset range of future time period airport periphery, the present solution provides a kind of machines based on LSTM model
Field traffic congestion prediction technique.Its method detailed process is as follows:
1. in real time obtain airport periphery preset range in road traffic related information and air station flight landing information and
Aeronautical meteorology information in the preset range of airport periphery；
2. the traffic related information, flight landing information and aeronautical meteorology information are inputted LSTM model；
3. obtaining the output of the LSTM model as a result, the output result is that model is preset on prediction future time period airport periphery
Enclose the congestion index of interior road.
The shortcomings that technology, is:
1. the accuracy of road prediction is not high, it is further contemplated that the factor of traffic condition is more influenced, as it is vaporous
Condition (technology only accounts for aeronautical meteorology factor), festivals or holidays factor and traffic accident etc..
2. since the input data of the input module of device includes flight landing information and aeronautical meteorology information, and airport
Road traffic prediction has particularity, i.e., is affected by aviation factor, therefore the universality of device is relatively low, is only applicable to machine
Field periphery traffic congestion prediction.
Summary of the invention
The purpose of the present invention is overcoming the shortcomings of existing methods, a kind of intelligent transportation status predication of machine learning is proposed
Method.The main problem that the present invention solves is how to pass through road net data collection, the magnitude of traffic flow-speed data collection, meteorological dataset
With social property data set, the Encoder-Decoder mixed model based on LSTM is established, by the mixed model
The intelligence prediction in real time of traffic behavior is realized in training, promotes the accuracy of each road traffic state prediction in each region.
To solve the above-mentioned problems, described the invention proposes a kind of intelligent transportation trend prediction method of machine learning
Method includes:
Acquire road net data collection, the magnitude of traffic flow-speed data collection, meteorological dataset, social property data set；
According to the magnitude of traffic flow-speed data collection, the Encoder-Decoder basic model based on LSTM is established；
According to social property data set and road net data collection, social property and road network category are added on the basic model
Property；
According to road net data collection, the magnitude of traffic flow-speed data collection and meteorological dataset, region is carried out using Voronoi diagram
It divides, calculates traffic and meteorological degree of correlation in each region, new meteorological attribute is generated with this.Add on the basic model
The meteorological attribute for adding generation, obtains mixed model；
According to the magnitude of traffic flow-speed data and meteorological data obtained in real time, according to described in the input of format as defined in model
Mixed model completes prediction in real time.
Preferably, described to generate new meteorological attribute, it specifically includes:
According to road net data collection, using Voronoi diagram division methods, urban area is divided into cell；
To each cell, construction does not include meteorological data and training dataset and test data comprising meteorological data
Collection；
To each cell, the MAPE value of model when not including meteorological data is calculated；
To each cell, the MAPE value of model when comprising meteorological data is calculated；
To each cell, the absolute difference of above-mentioned two MAPE value is calculated, new meteorological attribute is generated with this, is used for
It is inserted into the LSTM block of model.
A kind of intelligent transportation trend prediction method of machine learning proposed by the present invention, has fully considered time and space effect
It answers, and increases meteorological attribute, road network attribute, social property, be able to ascend the accuracy of road traffic prediction of speed；Wherein gas
As attribute generating process in, calculate on the basis of road junction in the cell region divided traffic speed and meteorological
Degree of correlation, it is contemplated that a possibility that traffic is not by meteorological effect in certain unit areas, therefore can be improved to all lists
The prediction accuracy in first region.
Detailed description of the invention
In order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, to embodiment or will show below
There is attached drawing needed in technical description to be briefly described, it should be apparent that, the accompanying drawings in the following description is only this
Some embodiments of invention for those of ordinary skill in the art without creative efforts, can be with
Other attached drawings are obtained according to these attached drawings.
Fig. 1 is the intelligent transportation trend prediction method overview flow chart of the embodiment of the present invention；
Fig. 2 is the basic model schematic diagram of the embodiment of the present invention；
Fig. 3 is the mixed model schematic diagram of the embodiment of the present invention.
Specific embodiment
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, complete
Site preparation description, it is clear that described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.It is based on
Embodiment in the present invention, it is obtained by those of ordinary skill in the art without making creative efforts every other
Embodiment shall fall within the protection scope of the present invention.
Fig. 1 is the intelligent transportation trend prediction method overview flow chart of the embodiment of the present invention, as shown in Figure 1, this method packet
It includes:
S1 acquires road net data collection, the magnitude of traffic flow-speed data collection, meteorological dataset, social property data set；
S2 establishes the Encoder-Decoder basic model based on LSTM according to the magnitude of traffic flow-speed data collection；
S3 adds social property and road network according to social property data set and road net data collection on the basic model
Attribute；
S4 carries out area using Voronoi diagram according to road net data collection, the magnitude of traffic flow-speed data collection and meteorological dataset
Domain divides, and calculates traffic and meteorological degree of correlation in each region, new meteorological attribute is generated with this.On the basic model
The meteorological attribute generated is added, mixed model is obtained；
S5 inputs institute according to format as defined in model according to the magnitude of traffic flow-speed data and meteorological data obtained in real time
Mixed model is stated, prediction in real time is completed.
Step S1, specific as follows:
S1-1: utilizing map software API, acquires the road network information in city, constructs road net data collection, and saves it in this
In ground storage device；
S1-2: the magnitude of traffic flow-speed data collection is acquired using map software API, and saves it in local storage
In；
S1-3: collecting the meteorological data of urban meteorological platform publication, constructs meteorological dataset, and save it in and be locally stored
In device；
S1-4: it from the open source data set such as Q-Traffic Dataset that relevant traffic data are analyzed, acquires society and belongs to
Property data set, and save it in local storage；
S1-5: for all data sets, being pre-processed, and invalid data will be rejected.Such as the magnitude of traffic flow-
Speed data is concentrated, and timestamp will be converted to the time of date format, calculates interval of time according to time and range data
Interior rolling average speed etc..
Step S2, specific as follows:
S2-1: the traffic speed variable { v before given1,v2,…,vt, establish the Encoder-Decoder based on LSTM
Basic model predicts following traffic speed variable { vt+1,vt+2,…,vt+t′}.It is now assumed that variable vtThere is k dimension, i.e.,LSTM model core is three kinds of threshold structures in storage unit --- forget door, input gate
And out gate, storage unit finally obtain state vector CtWith output vector ht；
S2-2: being made of formula (2-1) and forget door, by the way that connection matrix W is arrangedfTo determine from last moment output vector
ht-1Which information middle removal retains；
ft=σ (Wf·[ht-1,vt]+bf) (2-1)
Wherein, σ indicates activation primitive sigmod, vtFor dkDimension is originally inputted, ht-1For dhDimension output, hidden layer dimension are dc
(usual dh=dc), [ht-1,vt] be two vectors splicing, WfTo forget gate matrix, bfFor bias vector, through activation primitive by
F is obtained after point transformationt。
S2-3: constituting input gate by formula (2-2), determines and completes location mode vector CtUpdate；
Ct=ft*Ct-1+σ(Wf·[ht-1,vt]+bi)*tanh(Wc·[ht-1,vt]+bc) (2-2)
Wherein, Ct、Ct-1The respectively state vector of active cell state vector and a upper unit, σ, tanh are respectively indicated
Activation primitive sigmod and hyperbolic tangent function, vtIndicate dkDimension is originally inputted, ht-1For dhDimension output, [ht-1,vt] be two to
The splicing of amount, Wf、WcIt is to forget gate matrix and input gate matrix, b respectivelyi、bcFor bias vector.
S2-4: out gate is constituted by formula (2-3), to CtUsing tanh activation primitive, output vector h is obtainedt；
ht=σ (Wo·[ht-1,vt]+bo)*tanh(Ct) (2-3)
Wherein, CtFor active cell state vector, σ, tanh respectively indicate activation primitive sigmod and hyperbolic tangent function,
vtIndicate dkDimension is originally inputted, ht-1For dhDimension output, [ht-1,vt] be two vectors splicing, WoIt is output gate matrix, boIt is inclined
Set vector.
S2-5: by formula (2-4) by output vector htA full articulamentum (FC) is connected, predicted value v is finally obtainedt+1；
vt+1=σ (Wout·ht+bout) (2-4)
Wherein, σ indicates activation primitive sigmod, WoutTo connect layer matrix, h entirelytFor output vector, boutTo be biased towards
Amount, vt+1For the predicted value at t+1 moment.
S2-6: all traffic speed variable v are constituted according to the above methodi(i ∈ { 1,2 ..., t+t ' }) corresponding LSTM
Block, then by { v1,v2,…,vtLSTM block as the part Encoder, { vt+1,vt+2,…,vt+t′LSTM block conduct
The part Decoder, basic model are as shown in Figure 2.
Step S3, specific as follows:
S3-1: in the full articulamentum FC in each of the part basic model Decoder, addition social property is inputted, mainly
Including working peak period, the fields such as peak period, weekend, festivals or holidays of coming off duty；
S3-2: in the full articulamentum FC in each of the part basic model Decoder, addition road network attribute is inputted, mainly
Including road section ID, section title, road section length, initial position ID, initial position longitude and latitude, end position ID, end position warp
The fields such as latitude, number of track-lines, speed limit grade；
S3-3: obtaining a mid-module, and subsequent introduction for convenience is named as model temp.
Step S4, specific as follows:
Wherein, S4-1 to S4-5 is to generate new meteorological attribute:
S4-1: according to collected road net data collection, using Voronoi diagram division methods, urban area is divided into list
Member, division principle are the section that each cell contains at least one intersection and is connected with the intersection, wherein seed
It is road junction, the shapes and sizes of each cell are different；
S4-2: understanding for convenience, is given at the statement of the variable name to be used in subsequent introduction herein.It has been coupled a certain
The road net data in all sections and traffic-data on flows data set are named as data set A in cell in period.It is coupled
With the road net data, traffic-data on flows in all sections in cell in the data set A same period and meteorological data
Data set is named as data set B.Be coupled the road net data in all sections in cell in the A data set subsequent one period with
And traffic-data on flows data set is named as data set C；
S4-3: assuming that the traffic speed in data set C certain time period is { v1,v2,…,vk}.For each unit
Lattice, input data set A are trained the mid-module temp of S3, take out and the traffic speed in data set C at the same time section
Spend prediction result, it is assumed that beWherein a1,a2,…,akIndicate constant incremental since 1, i.e. a1=
1,ak=k.The MAPE value between the two sets of speeds is calculated, its calculation formula is formula (2-5), are as a result denoted as MAPEa；
S4-4: assuming that the traffic speed in data set C certain time period is { v1,v2,…,vk}.For each unit
Lattice, input data set B are trained mid-module temp, take out pre- with the traffic speed in data set C at the same time section
It surveys as a result, being assumed to beWherein b1,b2,…,bkIndicate constant incremental since 1, i.e. b1=1, bk
=k.The MAPE value between the two sets of speeds is calculated, its calculation formula is formula (2-6), are as a result denoted as MAPEb；
S4-5: the MAPE of each cell of division is acquired according to the above methodaAnd MAPEbValue, when the difference of the two values
When absolute value is greater than some threshold value (sets itself), then show after considering meteorologic factor, accuracy rate increases, it can be deduced that
Conclusion, the meteorological influence to the cell traffic is generally existing, therefore should add a column field attribute in meteorological attribute
It indicates that meteorological whether there is on the region influences, is named as weather_relation, its value is set to 1；Otherwise, then explanation exists
In this cell, meteorological influence be it is uncertain, should equally, in meteorological attribute add a column field attribute indicates meteorological right
The region is named as weather_relation, its value is set to 0 with the presence or absence of influence；
S4-6: in the input of each LSTM block of mid-module temp, newly-generated meteorological attribute input is added, mainly
Including fields such as current time, wind speed, wind direction, rainfall, Current Temperatures, weather_relation, mixed model is obtained.It is mixed
Molding type is as shown in figure 3, wherein AT_social indicates social property, and AT_road indicates section attribute, when AT_W (t) indicates t
The meteorological attribute at quarter.
Step S5, specific as follows:
S5-1: flow-speed data, the meteorological data that will be obtained in real time are pre-processed (in vain and repeat number such as removal
According to), and be associated according to the time；
S5-2: flow-speed data, the meteorological data after association are closed with road net data, social property data
And then input in mixed model, it is predicted in real time.
The embodiment of the present invention propose a kind of machine learning intelligent transportation trend prediction method, fully considered the time with
Three-dimensional effect, and meteorological attribute, road network attribute, social property are increased, it is able to ascend the accuracy of road traffic prediction of speed；
Wherein in the generating process of meteorological attribute, traffic speed in the cell region divided is calculated on the basis of road junction
With meteorological degree of correlation, it is contemplated that a possibility that traffic is not by meteorological effect in certain unit areas, therefore can be improved pair
The prediction accuracy of all unit areas.
Those of ordinary skill in the art will appreciate that all or part of the steps in the various methods of above-described embodiment is can
It is completed with instructing relevant hardware by program, which can be stored in a computer readable storage medium, storage
Medium may include: read-only memory (ROM, Read Only Memory), random access memory (RAM, Random
Access Memory), disk or CD etc..
In addition, the intelligent transportation trend prediction method for being provided for the embodiments of the invention a kind of machine learning above carries out
It is discussed in detail, used herein a specific example illustrates the principle and implementation of the invention, above embodiments
Explanation be merely used to help understand method and its core concept of the invention；At the same time, for those skilled in the art,
According to the thought of the present invention, there will be changes in the specific implementation manner and application range, in conclusion in this specification
Appearance should not be construed as limiting the invention.
Claims (2)
1. a kind of intelligent transportation trend prediction method of machine learning, which is characterized in that the described method includes:
Acquire road net data collection, the magnitude of traffic flow-speed data collection, meteorological dataset, social property data set；
According to the magnitude of traffic flow-speed data collection, the Encoder-Decoder basic model based on LSTM is established；
According to social property data set and road net data collection, social property and road network attribute are added on the basic model；
According to road net data collection, the magnitude of traffic flow-speed data collection and meteorological dataset, region division is carried out using Voronoi diagram,
Traffic and meteorological degree of correlation in each region are calculated, new meteorological attribute is generated with this.Life is added on the basic model
At meteorological attribute, obtain mixed model；
According to the magnitude of traffic flow-speed data and meteorological data obtained in real time, the mixing is inputted according to format as defined in model
Model completes prediction in real time.
2. a kind of intelligent transportation trend prediction method of machine learning as described in claim 1, which is characterized in that the generation
It the step of new meteorological attribute, specifically includes:
According to road net data collection, using Voronoi diagram division methods, urban area is divided into cell；
To each cell, construction does not include meteorological data and training dataset and test data set comprising meteorological data；
To each cell, the MAPE value of model when not including meteorological data is calculated；
To each cell, the MAPE value of model when comprising meteorological data is calculated；
To each cell, the absolute difference of above-mentioned two MAPE value is calculated, new meteorological attribute is generated with this, for being inserted into
The LSTM block of model.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910614671.XA CN110428613A (en) | 2019-07-09 | 2019-07-09 | A kind of intelligent transportation trend prediction method of machine learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910614671.XA CN110428613A (en) | 2019-07-09 | 2019-07-09 | A kind of intelligent transportation trend prediction method of machine learning |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110428613A true CN110428613A (en) | 2019-11-08 |
Family
ID=68409072
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910614671.XA Pending CN110428613A (en) | 2019-07-09 | 2019-07-09 | A kind of intelligent transportation trend prediction method of machine learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110428613A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113688687A (en) * | 2021-07-27 | 2021-11-23 | 中电海康集团有限公司 | Traffic jam state rapid identification and prediction method based on electric police data |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107610469A (en) * | 2017-10-13 | 2018-01-19 | 北京工业大学 | A kind of day dimension regional traffic index forecasting method for considering multifactor impact |
| CN108898838A (en) * | 2018-08-03 | 2018-11-27 | 首都经济贸易大学 | A kind of aerodrome traffic congestion prediction technique and device based on LSTM model |
| KR101951595B1 (en) * | 2018-05-18 | 2019-02-22 | 한양대학교 산학협력단 | Vehicle trajectory prediction system and method based on modular recurrent neural network architecture |
| CN109658695A (en) * | 2019-01-02 | 2019-04-19 | 华南理工大学 | A kind of multifactor Short-time Traffic Flow Forecasting Methods |
| CN109671272A (en) * | 2018-12-29 | 2019-04-23 | 广东工业大学 | A kind of freeway traffic flow prediction technique |
| CN109785629A (en) * | 2019-02-28 | 2019-05-21 | 北京交通大学 | A kind of short-term traffic flow forecast method |
| CN109886444A (en) * | 2018-12-03 | 2019-06-14 | 深圳市北斗智能科技有限公司 | A kind of traffic passenger flow forecasting, device, equipment and storage medium in short-term |
-
2019
- 2019-07-09 CN CN201910614671.XA patent/CN110428613A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107610469A (en) * | 2017-10-13 | 2018-01-19 | 北京工业大学 | A kind of day dimension regional traffic index forecasting method for considering multifactor impact |
| KR101951595B1 (en) * | 2018-05-18 | 2019-02-22 | 한양대학교 산학협력단 | Vehicle trajectory prediction system and method based on modular recurrent neural network architecture |
| CN108898838A (en) * | 2018-08-03 | 2018-11-27 | 首都经济贸易大学 | A kind of aerodrome traffic congestion prediction technique and device based on LSTM model |
| CN109886444A (en) * | 2018-12-03 | 2019-06-14 | 深圳市北斗智能科技有限公司 | A kind of traffic passenger flow forecasting, device, equipment and storage medium in short-term |
| CN109671272A (en) * | 2018-12-29 | 2019-04-23 | 广东工业大学 | A kind of freeway traffic flow prediction technique |
| CN109658695A (en) * | 2019-01-02 | 2019-04-19 | 华南理工大学 | A kind of multifactor Short-time Traffic Flow Forecasting Methods |
| CN109785629A (en) * | 2019-02-28 | 2019-05-21 | 北京交通大学 | A kind of short-term traffic flow forecast method |
Non-Patent Citations (1)
| Title |
|---|
| BINBING LIAO等: "Deep Sequence Learning with Auxiliary Information for Traffic Prediction", 《KDD’18：PROCEEDINGS OF THE 24TH ACM SIGKDD INTERNATIONAL CONFERENCE ON KNOWLEDGE DISCOVERY & DATA MING》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113688687A (en) * | 2021-07-27 | 2021-11-23 | 中电海康集团有限公司 | Traffic jam state rapid identification and prediction method based on electric police data |
| CN113688687B (en) * | 2021-07-27 | 2024-05-28 | 中电海康集团有限公司 | Traffic jam state rapid identification and prediction method based on electric warning data |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Liu et al. | Exploring the impact of connected and autonomous vehicles on freeway capacity using a revised Intelligent Driver Model | |
| CN109959388B (en) | Intelligent traffic refined path planning method based on grid expansion model | |
| CN101639871B (en) | Vehicle-borne dynamic traffic information induction system analog design method facing behavior research | |
| CN108898838A (en) | A kind of aerodrome traffic congestion prediction technique and device based on LSTM model | |
| CN109840660A (en) | A kind of vehicular characteristics data processing method and vehicle risk prediction model training method | |
| Ngoduy | Multiclass first-order modelling of traffic networks using discontinuous flow-density relationships | |
| CN105678034A (en) | Pedestrian crossing time model and signalized intersection pedestrian crosswalk width optimizing method | |
| CN104200127A (en) | Optimal path analyzing method based on road corner weight | |
| CN104821086B (en) | Method for positioning low-efficient road section combination in large-scale traffic network | |
| Park et al. | Calibration and validation of TRANSIMS microsimulator for an urban arterial network | |
| Lin et al. | Data-driven spatial-temporal analysis of highway traffic volume considering weather and festival impacts | |
| CN110428613A (en) | A kind of intelligent transportation trend prediction method of machine learning | |
| Cosido et al. | Automatic calculation of bicycle routes by combining meta-heuristics and GIS techniques within the framework of smart cities | |
| Liu et al. | Assessing the impacts of connected-and-autonomous vehicle management strategy on the environmental sustainability of urban expressway system | |
| Wang et al. | Dynamic traffic congestion simulation and dissipation control based on traffic flow theory model and neural network data calibration algorithm | |
| Li et al. | Optimization model and method of variable speed limit for urban expressway | |
| Song et al. | Impacts of connected and autonomous vehicles with level 2 automation on traffic efficiency and energy consumption | |
| Yang | RETRACTED ARTICLE: An intelligent method to select emergency allocation path of groundwater resource based on game theory | |
| Liu et al. | Research on urban transport network topology vulnerability identification under rainfall conditions | |
| Zhang et al. | Traffic Guidance Based on LSTM Neural Network and Dual Tracking Dijkstra Algorithm | |
| Song et al. | Visual methods for time-varying intersection traffic flow data | |
| Kong et al. | Findings on Queue Length Based Macroscopic Fundamental Diagrams with Enhanced Floating Car Estimation Method | |
| Hu et al. | City traffic forecasting using taxi GPS data: A coarse-grained cellular automata model | |
| Li et al. | Study on the reduced traffic congestion method based on dynamic guidance information | |
| Hao et al. | A day-to-day invariant macroscopic fundamental diagrams for probe vehicles |
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 |
Application publication date: 20191108 |
|
| RJ01 | Rejection of invention patent application after publication |