IEDSFAN: information enhancement and dynamic-static fusion attention network for traffic flow forecasting
Abstract Accurate forecasting of traffic flow in the future period is very important for planning traffic routes and alleviating traffic congestion. However, traffic flow forecasting still faces serious challenges. Most of the existing traffic flow forecasting methods are static graph convolutional...
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Springer
2024-11-01
|
| Series: | Complex & Intelligent Systems |
| Subjects: | |
| Online Access: | https://doi.org/10.1007/s40747-024-01663-1 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832571171535585280 |
|---|---|
| author | Lianfei Yu Ziling Wang Wenxi Yang Zhijian Qu Chongguang Ren |
| author_facet | Lianfei Yu Ziling Wang Wenxi Yang Zhijian Qu Chongguang Ren |
| author_sort | Lianfei Yu |
| collection | DOAJ |
| description | Abstract Accurate forecasting of traffic flow in the future period is very important for planning traffic routes and alleviating traffic congestion. However, traffic flow forecasting still faces serious challenges. Most of the existing traffic flow forecasting methods are static graph convolutional networks based on prior knowledge, ignoring the special spatial–temporal dynamics of spatial–temporal data. Using only adaptive dynamic graphs completely discards the objective and real spatial connectivity information in static graphs. To this end, we propose a novel information enhancement and dynamic-static fusion attention network (IEDSFAN). Firstly, the Multi-Graph Fusion Gating mechanism (MGFG) designed in IEDSFAN effectively fuses dynamic and static graphs to dynamically capture the hidden spatial–temporal correlation. Secondly, we construct a novel Gated Multi-head Self-Attention (GMHSA), which maps the input through the MGFG module to capture the complex spatial–temporal interactions in the features. Finally, we generate adaptive parameters to solve the problem that shared parameters cannot learn multiple traffic patterns, and enhance the expression of sequence information through the peak flag module. We conducted extensive experiments on five real-world traffic datasets, and the experimental results show that the performance of IEDSFAN is significantly better than all baselines. |
| format | Article |
| id | doaj-art-6e432fa04ff841b582643a7e2f79cf59 |
| institution | Kabale University |
| issn | 2199-4536 2198-6053 |
| language | English |
| publishDate | 2024-11-01 |
| publisher | Springer |
| record_format | Article |
| series | Complex & Intelligent Systems |
| spelling | doaj-art-6e432fa04ff841b582643a7e2f79cf592025-02-02T12:49:42ZengSpringerComplex & Intelligent Systems2199-45362198-60532024-11-0111111910.1007/s40747-024-01663-1IEDSFAN: information enhancement and dynamic-static fusion attention network for traffic flow forecastingLianfei Yu0Ziling Wang1Wenxi Yang2Zhijian Qu3Chongguang Ren4School of Computer Science and Technology, Shandong University of TechnologySchool of Computer Science and Technology, Shandong University of TechnologySchool of Computer Science and Technology, Shandong University of TechnologySchool of Computer Science and Technology, Shandong University of TechnologySchool of Computer Science and Technology, Shandong University of TechnologyAbstract Accurate forecasting of traffic flow in the future period is very important for planning traffic routes and alleviating traffic congestion. However, traffic flow forecasting still faces serious challenges. Most of the existing traffic flow forecasting methods are static graph convolutional networks based on prior knowledge, ignoring the special spatial–temporal dynamics of spatial–temporal data. Using only adaptive dynamic graphs completely discards the objective and real spatial connectivity information in static graphs. To this end, we propose a novel information enhancement and dynamic-static fusion attention network (IEDSFAN). Firstly, the Multi-Graph Fusion Gating mechanism (MGFG) designed in IEDSFAN effectively fuses dynamic and static graphs to dynamically capture the hidden spatial–temporal correlation. Secondly, we construct a novel Gated Multi-head Self-Attention (GMHSA), which maps the input through the MGFG module to capture the complex spatial–temporal interactions in the features. Finally, we generate adaptive parameters to solve the problem that shared parameters cannot learn multiple traffic patterns, and enhance the expression of sequence information through the peak flag module. We conducted extensive experiments on five real-world traffic datasets, and the experimental results show that the performance of IEDSFAN is significantly better than all baselines.https://doi.org/10.1007/s40747-024-01663-1Traffic flow forecastingSpatial–temporal correlationGraph convolutional networksInformation enhancementSelf-attention |
| spellingShingle | Lianfei Yu Ziling Wang Wenxi Yang Zhijian Qu Chongguang Ren IEDSFAN: information enhancement and dynamic-static fusion attention network for traffic flow forecasting Complex & Intelligent Systems Traffic flow forecasting Spatial–temporal correlation Graph convolutional networks Information enhancement Self-attention |
| title | IEDSFAN: information enhancement and dynamic-static fusion attention network for traffic flow forecasting |
| title_full | IEDSFAN: information enhancement and dynamic-static fusion attention network for traffic flow forecasting |
| title_fullStr | IEDSFAN: information enhancement and dynamic-static fusion attention network for traffic flow forecasting |
| title_full_unstemmed | IEDSFAN: information enhancement and dynamic-static fusion attention network for traffic flow forecasting |
| title_short | IEDSFAN: information enhancement and dynamic-static fusion attention network for traffic flow forecasting |
| title_sort | iedsfan information enhancement and dynamic static fusion attention network for traffic flow forecasting |
| topic | Traffic flow forecasting Spatial–temporal correlation Graph convolutional networks Information enhancement Self-attention |
| url | https://doi.org/10.1007/s40747-024-01663-1 |
| work_keys_str_mv | AT lianfeiyu iedsfaninformationenhancementanddynamicstaticfusionattentionnetworkfortrafficflowforecasting AT zilingwang iedsfaninformationenhancementanddynamicstaticfusionattentionnetworkfortrafficflowforecasting AT wenxiyang iedsfaninformationenhancementanddynamicstaticfusionattentionnetworkfortrafficflowforecasting AT zhijianqu iedsfaninformationenhancementanddynamicstaticfusionattentionnetworkfortrafficflowforecasting AT chongguangren iedsfaninformationenhancementanddynamicstaticfusionattentionnetworkfortrafficflowforecasting |