A Study on the Forecast Method of Urban Rail Transit

Abstract

Simple and quick short-term passenger flow forecast is the basis of improving the efficiency of urban rail transit operation. This paper discusses the advantages and disadvantages of BP neural network in the short-term passenger flow forecast of urban rail transit and puts forward to optimize the BP neural network forecast model by genetic algorithm method. After analyzing the actual data of time and space characteristics of passenger flow in urban rail transit, this paper determines the forecast object of the section passenger flow. Then, this paper selects the BP neural network as a forecast model and makes optimization or adjustment focus on the rate of convergence, local optimum, and overfitting for BP neural network. Making short-term section passenger flow forecast by using the passenger flow data and comparing the forecast results with the actual results, the model can be verified feasibility. Through this method, the optimized BP neural network can forecast passenger flow after 30 min accurately and provide the basis for transit operation.