Traffic Guidance Based on LSTM Neural Network and Dual Tracking Dijkstra Algorithm

Abstract

Based on the historical big data intelligent transportation system, the traffic map model is constructed according to the characteristics of the city. With the help of the LSTM (Long Short-Term Memory) neural network prediction function, the city environmental data and traffic data are fused and preprocessed. Through the optimization and function expansion of the shortest path double-tracking Dijkstra algorithm, a route guidance scheme to avoid traffic congestion and severe environmental pollution is provided. The system can be applied to actual road regulation and optimal path selection. Not only can solve the city's vehicle congestion, but also alleviate the problem of air pollution.