Temporal-Spatial Quantum Graph Convolutional Neural Network Based on Schrödinger Approach for Traffic Congestion Prediction

Traffic congestion prediction (TCP) plays a vital role in intelligent transportation systems due to its importance of traffic management. Methods for TCP have emerged greatly with the development of machine learning. However, TCP is always a challenging work due to the dynamic characteristics of traffic data and the complex structure of traffic network. This paper presents a new quantum algorithm that can capture temporal and spatial features of traffic data simultaneously for TCP. The algorithm consists of the following steps. First, the authors give a closed-form solution in the Schrödinger approach theoretically to analyze this TCP problem in time dimension. Then they can get the temporal features from the solution. At last, whey construct a quantum graph convolutional network and apply temporal features into it. Thus, the temporal-spatial quantum graph convolutional neural network is proposed. The feasibility of this method is proved through experiments on the simulation platform. The experimental results show the average error rate is 0.21 and can resist perturbation effectively.

Language

  • English

Media Info

Subject/Index Terms

Filing Info

  • Accession Number: 01901201
  • Record Type: Publication
  • Files: TRIS
  • Created Date: Dec 1 2023 9:13AM