Research on the Urban Bike-sharing Usage based on ARIMA Model

Abstract

This study forecasts the demand for shared bicycles using the ARIMA model, incorporating insights from extensive literature on urban bike-sharing usage. Through in-depth analysis, first-order differencing was identified as crucial for achieving stationarity, leading to the recommendation of the ARIMA (312) model. This model effectively encapsulates the dynamics of shared bicycle usage, evident from the significant ADF test results and the careful selection of model parameters based on ACF and PACF plots. The research reveals the complex relationship between urban mobility and shared bicycle systems, providing a comprehensive framework for predicting usage trends. These findings make a significant contribution to the discussion on sustainable urban transportation and offer practical guidance for city planners and bike-sharing operators to enhance service efficiency and meet user demands effectively. The precise prediction of shared bicycle demand by the ARIMA (3, 1, 2) model highlights the effectiveness of advanced time series analysis in understanding and predicting bike-sharing usage patterns. According to the predictions, there are clear seasonal fluctuations, with a cycle of four quarters during which the predicted values gradually increase. This study emphasizes the potential of shared bicycles to enhance urban mobility and points out the need for more regulated development and management strategies to address challenges posed by the rapid growth of the bike-sharing industry. By providing a detailed understanding of the factors influencing shared bicycle usage, this research contributes to optimizing bike-sharing systems, thereby aiding in the sustainability of urban transportation networks.