Prediction of gas concentration evolution with evolutionary attention-based temporal graph convolutional network

作者：

Highlights：

• Graph convolutional network is applied to capture the spatial dependence.

• Gated recurrent units is adopted to achieve the temporal dependence.

• Evolutionary attention is used to improve attention-based prediction performance.

• The combined model is developed to predict spatiotemporal evolution efficiently.

摘要

•Graph convolutional network is applied to capture the spatial dependence.•Gated recurrent units is adopted to achieve the temporal dependence.•Evolutionary attention is used to improve attention-based prediction performance.•The combined model is developed to predict spatiotemporal evolution efficiently.

论文关键词：Prediction of gas concentration,Graph convolutional network,Evolutionary attention,Spatiotemporal dependence,Gas distribution map

论文评审过程：Received 17 June 2021, Revised 24 December 2021, Accepted 17 March 2022, Available online 29 March 2022, Version of Record 5 April 2022.

论文官网地址：https://doi.org/10.1016/j.eswa.2022.116944