Metro Passenger Flow Prediction Based on Optimized BP Neural Network Algorithm

Abstract

In recent years, the metro passenger flow in big cities has been increasing, and some lines are often crowded and delayed, which brings great pressure to the metro operation and management department. Therefore, it is urgent to build a scientific and effective mathematical model, which can help the metro operation and management department to formulate a reasonable train scheduling plan. This paper provides a new algorithm called APSO. APSO is used to optimize the BP neural network, namely APSO-BP algorithm. Experiments show that APSO-BP has high accuracy for metro passenger flow prediction.