With the promotion of Rural Revitalization Strategy, county public transport ushered in a period of rapid development. Public transport vehicles are the most important means of public transportation for commuters in the county, but the overcrowding of commuter peak at noon greatly reduces the service level of public transport. In order to better guide the public transportation system to make operation plan in advance, achieve rapid scheduling and improve service level, it is necessary to select appropriate forecasting methods to predict the future trend of short-term passenger flow during the noon commute peak period. Based on the historical data of public transport passenger flow in Fugu County, this paper takes 10 minutes as the node, selects the data of the same period and adjacent period of adjacent weeks and days as the input data, uses Elman neural network and BP neural network model to forecast the passenger flow in peak period, and compares the prediction performance and accuracy of the two prediction methods.

CICTP