Forecasting of traffic flows at ferry terminals. A hybrid model

Abstract

This paper presents an agile approach for managing ferry terminals. We propose a double-stage model that forecasts traffic. The hybrid model is based on quantitative tools. First, we search for associations in data and then, we develop a time series analysis. The specific objectives of the study are firstly to survey and collect data at a disaggregated level in terms of passenger and vehicle flows and time periods; and secondly, to develop forecasts on short-term using a combination of the proposed techniques. We think that the simultaneous application of historical analogy and time series forecasting is a powerful tool to predict the future traffic. Additionally, the methodological procedure proposed in this paper is general enough to be applied to other traffics and ports.