The research of method of short-term traffic flow forecast based on GA-BP neural network and chaos theory

In order to improve the efficiency and precision of short-term traffic flow forecast, for the traditional BP neural network that has such shortcomings as slow convergent speed and easy convergence to the local minimum points, this paper adopts genetic algorithm to optimize BP network structure and sets up a forecast method combining chaos theory with neural network, and the methods of traffic flow forecast based on chaos theory and neural network theory are researched. According to the results of forecasting research of actual traffic flow data by adopting chaos BP neural network forecast method and chaos GA-BP neural network forecast method, it is effective to apply chaos neural network forecast algorithm to short-term traffic flow forecast and improved chaos GA-BP neural network forecast algorithm can improve the precision and timeliness of short-term traffic flow forecast.