Road section feature representation learning algorithm based on space-time diagram information maximization model

Abstract

The invention provides a road section feature representation learning algorithm based on a space-time diagram information maximization model. According to the method, the time performance of the roadsection state is considered, the time information of the road section is deeply mined, a maximized mutual information mechanism is adopted, and the mutual influence and interaction relationship amongthe road section information, the time information and the traffic information is extracted and utilized in a learning algorithm based on a neural network. The obtained road section representation better reflects the real-time global traffic condition, the real-time dependence relationship between the upstream and the downstream of the road section is learned, and thus greatly improving the travel time prediction precision.

Description

technical field [0001] The present invention relates to related fields such as graph neural networks, and more specifically, relates to a road segment feature representation learning algorithm based on a spatio-temporal graph information maximization model. Background technique [0002] With the rapid increase in the number of motor vehicles, urban traffic congestion is becoming more and more serious, which leads to a series of problems such as low travel efficiency and waste of resources. Travel time prediction plays a vital role in applications such as traffic management, route planning, carpooling, and vehicle dispatching. Nowadays, almost all travel service applications have this function, such as Google Maps, Baidu Maps, Didi, etc. With the support of accurate travel time estimates, users can reasonably plan their personal travel paths to avoid wasting time on congested roads. At the same time, cities can also provide reasonable route guidance to effectively alleviate...

AI Technical Summary

Problems solved by technology

However, the road network is a complex network, and this method still has the following problems on the road network, that is, 1) it will cause adjacent road sections to be indistinguishable

The goal of the methods proposed so far is to make the node representation and the adjacent node representation more similar, but in the actual road network, there will be a situation where only one of the two adjacent road segments is congested, so they should be more distinguishable; 2) it Does not consider the interaction of road segments and traffic conditions

The overall traffic conditions will affect each road segment in the road network, and the state of some key road segments will in turn determine the traffic condition; 3) it does not consider the time-varying nature of the road segment itself

Images