Prediction of urban traffic congestion time by BPneural network

Abstract

This paper studies the time prediction of traffic congestion, through the real-time speed of the car, The free flow speed, the morning and evening peaks, the number of weeks, the average speed of the car in this section are used as independent variables, and the time of traffic congestion is studied as a dependent variable. Through the MATLAB R2014a numerical simulation software for BP neural network operation, the neuron input layer is the real-time speed, free flow speed, morning and evening peak, week number, and average vehicle speed factor of the car after processing. The prediction result is good, which proves that the model is effective and reliable, and can estimate the time of the vehicle passing through the crowded road section.