Disaggregate Simulation Model for Travel Demand Forecasting

The conventional “four-step” forecasting model cannot be applied on behavioral characteristics of travelers. Considering traveler activities and the travel decision-making process, using a 24-h timescale, and adopting rule-based modeling and object-oriented programming, we constructed a disaggregate simulation model based on activities for the household/individual travel chain in this paper. We used the model to simulate the selection of daily activities/travel plans, including event/trip destination, location of departure/destination and event starting time, and duration to rationally predict the spatial and temporal distribution of travel demand. The model simulates the travel information of travelers within a 24-h period using 12,182 travel records. Categorization is based on trip purpose and activity starting time. The goodness of fit between the value predicted by the model and the actual value is high, which indicates that the model is objective, reliable, and capable of simulating the decision-making behaviors of travelers.