STP-TrellisNets+: Spatial-Temporal Parallel TrellisNets for Multi-Step Metro Station Passenger Flow Prediction

Abstract

The drastic increase of metro passengers in recent years inevitably causes the overcrowdedness in the metro systems. Accurately predicting passenger flows at metro stations is critical for efficient metro system management, which helps alleviate such overcrowdedness. Compared to the prevalent next-step prediction, multi-step passenger flow prediction could prominently increase the prediction duration and reveal finer-grained passenger flow variations, which better helps metro system management. Thus, in this paper, we address the problem of <i>multi-step metro station passenger (MSP) flow prediction</i>. In light of MSP flows&#x2019; unique spatial-temporal characteristics, we propose <i>STP-TrellisNets+</i>, which for the first time augments the newly-emerged temporal convolutional framework <i>TrellisNet</i> for multi-step MSP flow prediction. The temporal module of STP-TrellisNets+ (named <i>CP-TrellisNetsED</i>) employs a Closeness TrellisNet followed by a <i>Periodicity TrellisNets-based Encoder-Decoder (P-TrellisNetsED)</i> to jointly capture the short- and long-term temporal correlation of MSP flows. In parallel to CP-TrellisNetsED, its spatial module (named <i>GC-TrellisNetsED</i>) adopts a novel transfer flow-based metric to characterize the spatial correlation among MSP flows, and implements another TrellisNetsED on multiple <i>diffusion graph convolutional networks (DGCNs)</i> in time-series order to capture the dynamics of such spatial correlation. Extensive experiments with two large-scale real-world automated fare collection datasets demonstrate that STP-TrellisNets+ outperforms the state-of-the-art baselines.