CN115115094A - Traffic flow prediction method combining sequence local information and multi-sequence incidence relation - Google Patents
Traffic flow prediction method combining sequence local information and multi-sequence incidence relation Download PDFInfo
- Publication number
- CN115115094A CN115115094A CN202210588239.XA CN202210588239A CN115115094A CN 115115094 A CN115115094 A CN 115115094A CN 202210588239 A CN202210588239 A CN 202210588239A CN 115115094 A CN115115094 A CN 115115094A
- Authority
- CN
- China
- Prior art keywords
- sequence
- information
- traffic flow
- matrix
- 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.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 38
- 230000007246 mechanism Effects 0.000 claims abstract description 9
- 230000007787 long-term memory Effects 0.000 claims abstract description 7
- 230000006403 short-term memory Effects 0.000 claims abstract description 7
- 230000008569 process Effects 0.000 claims abstract description 4
- 239000011159 matrix material Substances 0.000 claims description 41
- 239000013598 vector Substances 0.000 claims description 13
- 230000004927 fusion Effects 0.000 claims description 5
- 230000004913 activation Effects 0.000 claims description 3
- 230000006870 function Effects 0.000 claims description 3
- 238000004364 calculation method Methods 0.000 claims description 2
- 230000008859 change Effects 0.000 abstract description 6
- 230000003068 static effect Effects 0.000 abstract description 5
- 238000010801 machine learning Methods 0.000 description 3
- 238000013528 artificial neural network Methods 0.000 description 2
- 238000013527 convolutional neural network Methods 0.000 description 2
- 238000013135 deep learning Methods 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 238000007726 management method Methods 0.000 description 2
- 238000005065 mining Methods 0.000 description 2
- 238000011144 upstream manufacturing Methods 0.000 description 2
- 206010039203 Road traffic accident Diseases 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000011217 control strategy Methods 0.000 description 1
- 238000007418 data mining Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000007499 fusion processing Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- YHXISWVBGDMDLQ-UHFFFAOYSA-N moclobemide Chemical compound C1=CC(Cl)=CC=C1C(=O)NCCN1CCOCC1 YHXISWVBGDMDLQ-UHFFFAOYSA-N 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000012549 training Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/26—Government or public services
Landscapes
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Strategic Management (AREA)
- Human Resources & Organizations (AREA)
- Tourism & Hospitality (AREA)
- Economics (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- Marketing (AREA)
- Theoretical Computer Science (AREA)
- Development Economics (AREA)
- General Physics & Mathematics (AREA)
- Educational Administration (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Game Theory and Decision Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Traffic Control Systems (AREA)
Abstract
The invention discloses a traffic flow prediction method combining sequence local information and a multi-sequence incidence relation. The method comprises the steps of firstly extracting local dynamic change information of a historical flow sequence through a long-term and short-term memory network, then extracting global incidence relation information among different sequences through a graph convolution network, and finally combining the local dynamic change information with multi-sequence incidence relation information through a gating mechanism. In the process of extracting the global relationship information, the dynamic relationship of different sequences changing along with time is further considered on the basis of the static association relationship so as to realize more accurate traffic flow prediction.
Description
Technical Field
The invention belongs to the field of data mining and intelligent traffic, and particularly relates to a traffic flow prediction method combining sequence local information and a multi-sequence incidence relation.
Background
With the improvement of living standards and the development of urban traffic, a series of traffic problems are brought by rapidly increasing travel demands, and problems such as road congestion, traffic accidents and the like are frequent. In this context, establishing an effective intelligent transportation system is an important content for assisting transportation institutions to make scientific management decisions, wherein how to realize accurate prediction of traffic flow is an important component of the intelligent transportation system. Accurate traffic flow prediction can assist in making a real-time control strategy, and has important significance for scientific traffic management planning and safe and efficient travel of residents.
Traffic flow prediction refers to predicting future traffic flow by analytical mining of historical flow data. At present, the method for predicting the traffic flow by researchers at home and abroad can be mainly divided into a statistical learning-based method, a traditional machine learning-based method and a deep learning-based method.
A typical representative of the statistical learning-based method is a History Average (HA) method, which calculates an Average value of historical contemporaneous flows as a current predicted value, but is not suitable for dynamically changing traffic flow data; differential mean Moving AutoRegressive (ARIMA) is another typical statistical-based method that predicts unstable sequences by differentially converting them to stationary sequences; however, the effectiveness of the above-described statistical learning method-based prediction model depends heavily on the data quality.
Because the traffic flow has significant characteristics of nonlinearity and uncertainty, the machine learning method is also widely applied to traffic flow prediction, such as K Nearest Neighbors (KNN) algorithm, bayesian model, Support Vector Regression (SVR) algorithm, and the like. Although the machine learning method can model the nonlinear relationship between data, the requirement for the data features is high, and complicated feature processing is often required.
Because the deep neural network can effectively model a high-dimensional space-time data relationship and does not need complex artificial feature engineering, deep learning is widely applied to traffic flow prediction in recent years. For example, Long and Short-term Memory Network (LSTM) methods consider the dependency relationship between flows, and Convolutional Neural Network (CNN) methods preliminarily consider the spatio-temporal dependency relationship between different flow data, but the modeling of the spatio-temporal dependency relationship is rough, and it is difficult to realize effective expression of traffic Network data. Graph Neural Network (GNN) can realize effective modeling of flow data space-time characteristics based on road Network data by virtue of its strong non-european space modeling capability, and has become one of the main means for traffic flow prediction at present, and Graph Convolution Network (GCN) in GNN is mainly used for traffic flow prediction at present.
However, the current traffic flow prediction method generally performs overall modeling on different flow sequences to consider the global relationship between the sequences, or only considers the dynamic change of the flow sequences, and lacks the work of combining the two and mining the potential information contained in the flow data to improve the traffic flow prediction accuracy; in addition, the static graph depended by the existing GCN-based method is difficult to accurately reflect the dynamic correlation relation between different nodes along with the change of time.
Disclosure of Invention
In view of the problem that the traffic flow information extracted by the existing traffic flow prediction work is difficult to effectively utilize the complex mode and the dynamic association relation in the data, the invention provides a traffic flow prediction method combining sequence local information and multi-sequence association relation by combining the dynamic change information of a flow history sequence and the global relation information among a plurality of flow sequences, and realizes accurate traffic flow prediction by considering the dynamic relation of different sequences changing along with time in the extraction process of the global relation information.
The method comprises the following specific steps:
and (1) acquiring historical traffic flow of the N flow sensors at T moments to form a traffic flow data set.
And (2) forming a directed road network graph G (S, E) according to the geographic positions of the sensors, wherein S is a vertex set in the directed graph, each vertex represents one flow sensor, and E is a set of directed edges in the directed graph.
Extracting sequence local information, inputting historical flow data into a Long and Short-term Memory Network (LSTM) to obtain sequence local information C L 。
Step (4), extracting the multi-sequence incidence relation information, comprising the following substeps:
step (4.1) determining a geographic position adjacency matrix A between different sensors based on the road network G (S, E) G ；
Step (4.2) forming a historical flow correlation matrix A according to the historical flow data of each node on the graph T ；
Step (4.3) adjacency of geographic position to matrix A G And historical traffic adjacency matrix A T Combine to form a dynamic adjacency matrix A D ；
Step (4.4) dynamic adjacency matrix A D Inputting historical flow data into a Graph Convolution Network (GCN) to obtain multi-sequence association relation information C G ；
Step (5) extracting sequence local information C based on LSTM L And multi-sequence incidence relation information C extracted based on GCN G Performing information fusion based on a gating mechanism to obtain a final information representation C;
step (6), multi-scale traffic flow prediction comprises the following sub-steps:
step (6.1), determining the number K of moments to be subjected to flow prediction, and selecting K1 multiplied by 1 convolution kernels based on the number K;
and (6.2) applying K different convolution kernels to C to obtain K time traffic flow prediction results.
The invention has the following beneficial effects: in the extraction process of the multi-sequence incidence relation, considering that most of the existing GCN-based methods only model a single static graph, the invention combines a historical traffic relation correlation matrix which dynamically changes along with time to form a dynamic adjacency matrix on the basis of the traditional geographic position-based adjacency matrix, and constructs a dynamic sequence relation graph through the dynamic adjacency matrix to realize more accurate and effective traffic flow prediction.
Drawings
FIG. 1 is a flow chart of the method of the present invention;
FIG. 2 is a schematic diagram of an LSTM.
Detailed Description
The invention designs a traffic flow prediction method combining sequence local information and multi-sequence incidence relation aiming at the defect that the current flow prediction method only considers the global relation among different sequences (nodes and sensors) and considers the global relation as a static relation or only considers the dynamic change of the flow sequence.
The following will specifically describe a traffic flow prediction method based on sequence-sequence relation modeling, and the execution process of the method is shown in fig. 1.
For convenience of description, the associated symbols are defined as follows:
flow matrix
And (4) forming a flow matrix by historical traffic flow of the N sensors at the T time points.
The road network G (S, E) is a directed graph, wherein S is a vertex set in the directed graph, and a vertex v belongs to S and represents different sensors. And E is a set of directed edges in the directed graph, the directed edge E belongs to E, and when two sensors are within a certain distance threshold value d and a relationship between an upstream sensor and a downstream sensor exists, one directed edge is generated by connecting the upstream sensor with the downstream sensor.
At time t, all sensingVector of current time flow of device
Represents one column of the traffic matrix V, where
Indicating the flow at the ith sensor at time t.
The flow sequence formed by the sensor i at all times is
And one row in a flow matrix V is shown, a section with the length of w is intercepted during each training, and a matrix Q formed by the flow intercepted by all the sensors is one of the submatrices of V.
The method comprises the following specific steps:
step (1), data acquisition: and acquiring historical traffic flow of T time points of N sensors to form a historical flow matrix V.
And (2) forming a road network G (S, E) according to the geographical position of the sensor, and setting a time window as w.
Step (3) extracting sequence local information, and acquiring the flow x of all the sensors at w time steps according to the time window w t-w+1 ,…,x t . As shown in FIG. 2, the current flow of all sensors at each time is constructed into a vector x t Sequentially inputting Long and Short-term Memory Network (LSTM) units, and combining the LSTM unit with the hidden vector h of the previous moment t-1 And cell state vector c t-1 Obtaining a hidden vector h output at the current moment t And cell state vector c t . The whole process comprises the following calculation steps:
i t ＝σ(W xi x t +W hi h t-1 +b i )#(3.1)
f t ＝σ(W xf x t +W hf h t-1 +b f )#(3.2)
o t ＝σ(W xo x t +W ho h t-1 +b o )#(3.3)
c t ＝f t ⊙c t-1 +i t ⊙tanh(W xc x t +W hc h t-1 +b c )#(3.4)
h t ＝o t ⊙tanh(c t )#(3.5)
wherein i t ，f t ，o t An input gate, a forgetting gate and an output gate at time t, W represents a learnable parameter matrix, b i ，b f ，b o ，b c Denotes an offset vector, σ denotes a sigmoid activation function, tanh denotes a hyperbolic tangent activation function, and |, denotes a multiplication of parity elements.
The hidden vector h output at all the time instants t Splicing as extracted sequence local information C L 。
Step (4) extracting multi-sequence incidence relation information
Step (4.1) obtaining a geographic position neighbor matrix on the basis of the directed graph G (S, E)
Specifically, if there is a directed edge between sensor i and sensor j pointing from i to j, the element adjacent to the ith row and jth column in the matrix is A G [i][j]Set to 1, otherwise set to 0.
Step (4.2) modeling dynamic historical traffic flow related matrix A based on historical traffic flow data T . Specifically, A T [i][j]Representing the historical flow similarity between sensor i and sensor j by calculating
Wherein
Indicates the ith sensorIntercepting the average flow of the historical flow sequence, A T [i][j]The closer to 1 the absolute value of (a) indicates the higher the historical flow similarity between sensor i and sensor j.
And (4.3) the conventional traffic prediction method based on the Graph volume Network (GCN) generally utilizes the distance between nodes to establish a static Graph, but ignores the condition that the correlation between the nodes at different time points changes. The invention combines the dynamic historical traffic flow related matrix established based on historical traffic flow data of different nodes to form different dynamic adjacent matrixes at different moments on the basis of the adjacent matrix based on distance
A D ＝αA G +βA T #(4.2)
Wherein α, β represent weight coefficients. In particular, due to A G An element of 0 above indicates that the two nodes are not adjacent in distance or there is no flow direction relationship for traffic, so for A G The element with 0 above, in the final dynamic adjacency matrix A D Is also kept at 0 at the corresponding position.
Step (4.4) of enabling the dynamic adjacent matrix A D Inputting the historical flow vector q of each node into GCN to obtain multi-sequence incidence relation information C G (ii) a In the GCN, the update formula of the node is:
wherein
A matrix of the degree of representation,
the matrix of the unit is expressed by,
a matrix formed by all node information of the l-th layer is represented,
representing a parameter matrix learnable at layer l, in particular, H (0) ＝Q。
The whole picture is updated for L times to obtain H (L) Representing the updated information set of all nodes, namely the final sequence incidence relation information C G 。
Step (5) extracting sequence local information C based on LSTM L And multi-sequence incidence relation information C extracted based on GCN G And performing information fusion based on a gating mechanism to obtain a final information representation C. The information fusion process comprises the following steps:
first, the weight g is obtained through a gating mechanism c ：
g c ＝σ(W G C G +W L C L +b c )#(4.4)
Wherein W G ，
For learnable parameter matrices in the gating mechanism, b c Is a bias vector.
Then according to the learned weight g c Sequence local information C L And multi-sequence association relation information C G Fusion to obtain the final information representation C:
C＝g c ⊙C G +(1-g c )⊙C L #(4.5)
and (6) after the final information representation C is obtained, multi-scale traffic flow prediction is carried out according to the requirement.
Step (6.1) assuming that the current time is t and the target is to predict the traffic flow at K times after the t time, K different 1 × 1 convolution kernels K are selected 1 ，…，k K ；
And (6.2) applying K different convolution kernels to C to obtain traffic flow prediction results at K moments in the future.
Claims (7)
1. The traffic flow prediction method combining the sequence local information and the multi-sequence incidence relation is characterized by comprising the following specific steps:
acquiring historical traffic flow of N flow sensors at T moments to form a traffic flow data set;
step (2), forming a directed road network graph G (S, E) according to the geographic positions of the sensors, wherein S is a vertex set in the directed road network graph, each vertex represents a flow sensor, and E is a set of directed edges in the directed road network graph;
extracting sequence local information, inputting historical flow data into a long-term and short-term memory network to obtain sequence local information C L ；
And (4) extracting multi-sequence incidence relation information, comprising the following substeps:
step (4.1) determining a geographic position adjacency matrix A between different sensors based on the road network G (S, E) G ；
Step (4.2) forming a historical flow correlation matrix A according to the historical flow data of each node on the graph T ；
Step (4.3) adjacency of geographic position to matrix A G And historical traffic adjacency matrix A T Combine to form a dynamic adjacency matrix A D ；
Step (4.4) dynamic adjacency matrix A D Inputting historical flow data into a graph convolution network to obtain multi-sequence incidence relation information C G ；
Step (5) sequence local information C L And multi-sequence association relation information C G Performing information fusion based on a gating mechanism to obtain a final information representation C;
step (6), multi-scale traffic flow prediction comprises the following sub-steps:
step (6.1), determining the number K of moments to be subjected to flow prediction, and selecting K1 multiplied by 1 convolution kernels based on the number K;
and (6.2) applying K different convolution kernels to C to obtain K time traffic flow prediction results.
2. The traffic flow prediction method according to claim 1, wherein the step (3) is specifically:
acquiring the flow of all sensors at w time steps according to the time window w;
a vector x formed by the current time flow of all the sensors at each time t Inputting the data into the long-term and short-term memory networks in sequence;
splicing hidden vectors output by the long-term and short-term memory network at all times to obtain extracted sequence local information C L 。
3. The traffic flow prediction method according to claim 1, characterized in that the step (4.1) is specifically:
on the basis of the directed graph G (S, E), obtaining a geographic position neighbor matrix A G If a directed edge pointing from i to j exists between sensor i and sensor j, the element adjacent to the ith row and jth column in the matrix is A G [i][j]Set to 1, otherwise set to 0.
4. A traffic flow prediction method according to claim 3, characterised in that step (4.2) is specifically: with A T [i][j]Representing the historical flow similarity between sensor i and sensor j, then:
wherein
Indicating that the ith sensor intercepts the average flow of the historical flow sequence,
representing the flow at the ith sensor at time ti, and w is the time window.
5. The traffic flow prediction method according to claim 4, characterized in that the step (4.3) is specifically: dynamic adjacency matrix A D The calculation is as follows:
A D ＝αA G +βA T
wherein α, β represent weight coefficients.
6. The traffic flow prediction method according to claim 5, characterized in that in step (4.4), the update formula of the nodes in the graph convolution network is:
wherein D represents a degree matrix, I N Represents a unit matrix, H (l) A matrix W representing information of all nodes of the l-th layer (l) Representing a parameter matrix which can be learnt by the l-th layer, wherein ReLU represents an activation function;
the whole picture is updated for L times to obtain H (l) Representing the updated information set of all nodes, namely the final sequence incidence relation information C G 。
7. The traffic flow prediction method according to any one of claims 1 to 6, wherein in the step (5), the process of fusing information by a gating mechanism is:
first, the weight g is obtained through a gating mechanism c ：
g c ＝σ(W G C G +W L C L +b c )
Wherein W G ,W L For learnable parameter matrices in the gating mechanism, b c Is a bias vector;
then according to the weight g c Sequence partial information C L And multi-sequence association relation information C G Fusion to obtain the final information representation C:
C＝g c ⊙C G +(1-g c )⊙C L 。
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210588239.XA CN115115094B (en) | 2022-05-26 | 2022-05-26 | Traffic flow prediction method combining sequence local information and multi-sequence association relation |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210588239.XA CN115115094B (en) | 2022-05-26 | 2022-05-26 | Traffic flow prediction method combining sequence local information and multi-sequence association relation |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115115094A true CN115115094A (en) | 2022-09-27 |
| CN115115094B CN115115094B (en) | 2024-02-06 |
Family
ID=83326655
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210588239.XA Active CN115115094B (en) | 2022-05-26 | 2022-05-26 | Traffic flow prediction method combining sequence local information and multi-sequence association relation |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115115094B (en) |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070208498A1 (en) * | 2006-03-03 | 2007-09-06 | Inrix, Inc. | Displaying road traffic condition information and user controls |
| CN112418547A (en) * | 2020-12-03 | 2021-02-26 | 北京工业大学 | Bus stop passenger flow prediction method based on GCN-LSTM combined model |
| US20210133569A1 (en) * | 2019-11-04 | 2021-05-06 | Tsinghua University | Methods, computing devices, and storage media for predicting traffic matrix |
| CN113268916A (en) * | 2021-04-07 | 2021-08-17 | 浙江工业大学 | Traffic accident prediction method based on space-time graph convolutional network |
| CN113971479A (en) * | 2021-10-01 | 2022-01-25 | 浙江省机电设计研究院有限公司 | Daily traffic flow parameter prediction method integrating holiday and festival effects |
| US11238729B1 (en) * | 2020-09-11 | 2022-02-01 | Toyota Motor Engineering & Manufacturing North America, Inc. | Systems and methods for traffic flow prediction |
| CN114169647A (en) * | 2022-01-07 | 2022-03-11 | 重庆大学 | Traffic prediction method and system of continuous memory self-adaptive heterogeneous space-time diagram convolution |
| CN114202122A (en) * | 2021-12-13 | 2022-03-18 | 河北师范大学 | Urban traffic flow prediction method based on Markov cluster map attention network |
-
2022
- 2022-05-26 CN CN202210588239.XA patent/CN115115094B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070208498A1 (en) * | 2006-03-03 | 2007-09-06 | Inrix, Inc. | Displaying road traffic condition information and user controls |
| US20210133569A1 (en) * | 2019-11-04 | 2021-05-06 | Tsinghua University | Methods, computing devices, and storage media for predicting traffic matrix |
| US11238729B1 (en) * | 2020-09-11 | 2022-02-01 | Toyota Motor Engineering & Manufacturing North America, Inc. | Systems and methods for traffic flow prediction |
| CN112418547A (en) * | 2020-12-03 | 2021-02-26 | 北京工业大学 | Bus stop passenger flow prediction method based on GCN-LSTM combined model |
| CN113268916A (en) * | 2021-04-07 | 2021-08-17 | 浙江工业大学 | Traffic accident prediction method based on space-time graph convolutional network |
| CN113971479A (en) * | 2021-10-01 | 2022-01-25 | 浙江省机电设计研究院有限公司 | Daily traffic flow parameter prediction method integrating holiday and festival effects |
| CN114202122A (en) * | 2021-12-13 | 2022-03-18 | 河北师范大学 | Urban traffic flow prediction method based on Markov cluster map attention network |
| CN114169647A (en) * | 2022-01-07 | 2022-03-11 | 重庆大学 | Traffic prediction method and system of continuous memory self-adaptive heterogeneous space-time diagram convolution |
Non-Patent Citations (3)
| Title |
|---|
| BING YU 等: "Spatio-Temporal Graph Convolutional Networks: A Deep Learning Framework for Traffic Forecasting", HTTPS://ARXIV.ORG/PDF/ARXIV:1709.04875V4, pages 1 - 7 * |
| STRANGER_SS: "一种融合局部和全局空间相关性的交通预测混合模型", HTTPS://BLOG.CSDN.NET/STRANGER_SS/ARTICLE/DETAILS/124716999, pages 2 - 10 * |
| 高小婷;: "基于时空相关性的短时交通流预测模型", 电脑知识与技术, no. 18 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115115094B (en) | 2024-02-06 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111612243B (en) | Traffic speed prediction method, system and storage medium | |
| CN110223517B (en) | Short-term traffic flow prediction method based on space-time correlation | |
| CN112418547A (en) | Bus stop passenger flow prediction method based on GCN-LSTM combined model | |
| CN111079931A (en) | State space probabilistic multi-time-series prediction method based on graph neural network | |
| CN113268916A (en) | Traffic accident prediction method based on space-time graph convolutional network | |
| CN112270355B (en) | Active safety prediction method based on big data technology and SAE-GRU | |
| CN110675632B (en) | Vehicle short-time trajectory prediction control method aiming at multi-feature space and data sparseness | |
| CN115376317B (en) | Traffic flow prediction method based on dynamic graph convolution and time sequence convolution network | |
| CN113628249A (en) | RGBT target tracking method based on cross-modal attention mechanism and twin structure | |
| CN113538910A (en) | Self-adaptive full-chain urban area network signal control optimization method | |
| CN113705877A (en) | Real-time monthly runoff forecasting method based on deep learning model | |
| Barzola-Monteses et al. | Energy consumption of a building by using long short-term memory network: a forecasting study | |
| CN114169647A (en) | Traffic prediction method and system of continuous memory self-adaptive heterogeneous space-time diagram convolution | |
| CN113112791A (en) | Traffic flow prediction method based on sliding window long-and-short term memory network | |
| CN113487856A (en) | Traffic flow combination prediction model based on graph convolution network and attention mechanism | |
| CN112766603A (en) | Traffic flow prediction method, system, computer device and storage medium | |
| CN115376318B (en) | Traffic data compensation method based on multi-attribute fusion neural network | |
| CN116935649A (en) | Urban traffic flow prediction method for multi-view fusion space-time dynamic graph convolution network | |
| CN115755219A (en) | Flood forecast error real-time correction method and system based on STGCN | |
| CN112927507B (en) | Traffic flow prediction method based on LSTM-Attention | |
| CN113128666A (en) | Mo-S-LSTMs model-based time series multi-step prediction method | |
| CN117494034A (en) | Air quality prediction method based on traffic congestion index and multi-source data fusion | |
| Samal et al. | Time series forecasting of air pollution using deep neural network with multi-output learning | |
| CN115115094A (en) | Traffic flow prediction method combining sequence local information and multi-sequence incidence relation | |
| Yu et al. | An incremental learning based convolutional neural network model for large-scale and short-term traffic flow |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |