Traffic speed prediction method based on time sequence diagram neural network

Abstract

The invention discloses a traffic speed prediction method based on a time sequence diagram neural network, and the method comprises the following steps: S1, collecting the observation data of a traffic speed sensor network, and constructing a traffic map and a speed observation sequence; S2, enabling the coding end to carry out feature transformation on original node features; S3, carrying out node space feature fusion; S4, modeling dynamic time sequence characteristics in the network based on the bidirectional time sequence space coding layer; S5, enabling the decoding end to perform feature transformation on the original node features; S6, learning coding time sequence features of the splicing features based on a bidirectional GRU layer; and S7, based on the timing sequence multi-head attention layer, calculating the attention between the state of the current moment and a plurality of observation states of the coding end, and performing prediction. According to the method, the modeling problem of the time sequence characteristics and the spatial characteristics in the time sequence traffic network of the static topology is solved, the capturing capability of the traffic speed prediction model on the spatial characteristics and the time sequence dependence characteristics is improved based on the time sequence diagram neural network, and the method has good availability.

Description

technical field [0001] The invention belongs to the technical field of traffic network information, and in particular relates to a traffic speed prediction method based on a sequence diagram neural network. Background technique [0002] In many cities, especially large cities in developing countries, the existing transportation system is increasingly overwhelmed by the rapidly expanding demand for private cars and travel. Traffic congestion, long commutes and traffic accidents have seriously affected people. The standard of living and urban development, therefore, the construction of intelligent transportation system is getting more and more attention. Traffic speed prediction is an important part of smart transportation. The prediction on the traffic network has the following challenges: [0003] (1) Complex spatial dependencies: Roads at different locations in the traffic network have different transportation pressures. For example, on a main road, the downstream roads ar...

AI Technical Summary

Problems solved by technology

In addition, traffic congestion and other problems are more likely to occur in dense residential areas or urban traffic networks in office and residential areas than in other places

[0004] (2) The time dependence of dynamic changes: at...

Images