Online car-hailing order demand prediction method based on space-time context attention network

Abstract

In order to more reasonably allocate online car-hailing resources, reduce the empty driving rate of online car-hailing and improve the operation efficiency of online car-hailing, the invention provides an online car-hailing order demand prediction method based on a space-time context attention network, a network model constructed in the method comprises stacked space-time blocks and an output block, and each space-time block comprises TRELLIS-GRU and GAT, the TRELLIS-GRU is a gated cyclic grid network, the GAT is a graph attention network, the TRELLIS-GRU layer is used for fusing spatial and temporal features, and the GAT layer is used for mining spatial dependence in different time slices. According to the invention, influence of the spatial position of the urban area, the mutual relationship between different areas and the time dependence of historical orders in multiple time periods on the online car-hailing order demand is fully considered, so that the accuracy of the online car-hailing order demand in the urban area is improved.

Description

technical field [0001] The invention relates to the interdisciplinary technical fields of urban transportation planning and deep learning, and in particular to a demand prediction method for online car-hailing orders based on a spatio-temporal context attention network. Background technique [0002] With the rapid economic development and the rise of emerging travel methods such as Didi Chuxing and Dongfeng Travel, online car-hailing has become an important way for the public to travel. As of 2019, the proportion of online car-hailing passenger traffic to the total taxi passenger traffic It has reached 36.3%. The current online car-hailing service still has some problems of inefficient operation, such as long waiting time for passengers and high number of empty-loaded vehicles. Contents of the invention [0003] The present invention proposes a network-based car-hailing order demand prediction method based on a spatio-temporal context attention network, which is used to s...

AI Technical Summary

Problems solved by technology

[0003] The present invention proposes a network-based car-hailing order demand prediction method based on spatio-temporal context attention network, which is used to solve or at least partly solve the technical problem that existing methods cannot accurately predict network car-hailing order demand

Images