Four-step planning models commonly represent travel demand as an aggregate set of zonal productions and attractions. While this modelling paradigm is quite satisfactory for typical strategic planning and local impact studies, there are occasions where modellers need to know not only how many people are travelling, but who and why they are travelling. This paper shows how household travel survey data can be integrated into a point-to-point traffic assignment technique to maintain demographic and trip purpose information throughout the modelling process. The technique is embodied in a GIS-based model, called "RoadLink". The RoadLink model is used with data from the Victorian Activity and Travel Survey (VATS) to model traffic flows in Melbourne and shown to be a reasonable predictor of general travel patterns. The model is applied in a brief case study to demonstrate the rich demographic outputs potentially available from a disaggregate assignment process. (a).
Abstract