Short-term Passenger Flow Prediction on Bus Stop Based on Hybrid Model

Abstract

Abstract-Short-term passenger flow prediction on bus stop is an important base and technical support for bus dispatch strategy. In this paper, a new hybrid prediction model including two single models of BP neural network and time series model was proposed according to the periodicity and randomness properties of short-term passenger flow. By using the IC card data from buses, a three-layer BP neural network model was established to reflect the characteristics of the stability of shortterm passenger flow cycle. The time series model of cash ticket passenger flow data was established. Finally, the results of the two models were fitted to get the final prediction results. The results show that the proposed hybrid prediction model can effectively predict short-term passenger flow on bus stop.