Modelling Bicycle Demand Using Autoregressive and Moving Average Models

摘要

Bicycle transport is gaining importance in modern cities,therefore the modeling of its demand is most needed for sustainable city development.In this paper the measurement of number of bicyclists at one of the Czech Republic street was conducted.Monthly data in hourly periods was used.The autoregressive and moving average models of bicycle demand was calculated.Depending on ARMA parameters,to estimate the relative quality of statistical models,the Akaike Information Criterion(AIC)was used for a given set of data.In order to verify the correctness of the ARMA model,the residual autocorrelation analysis was carried out.The verification of the model's fit to the real data was made using the correlation coefficient.