Personal Activity Travel and Location Allocation Simulation
HBA has developed and implemented a particular form of advanced disaggregate tour-based personal travel model system, called the Personal Activity Travel and Location Allocation Simulation, or ‘PATLAS’, in a number of jurisdictions, including the State of California, the San Joaquin Valley, and the Calgary Region (jointly with City of Calgary staff).
The PATLAS model is a tour-based micro-simulation, where the individual trips in the tours made by each person in the (synthesized) population are represented explicitly and combined to establish trip tables that are loaded to a set of multi-modal networks. It considers the underlying activity patterns of individual model region residents as the key to travel decisions, and it uses the concept of a “tour” as a unit of analysis in the development of model components. A tour represents a closed or half closed chains of trips starting and ending at home or at the workplace. Each tour includes at least one destination and at least two successive trips. A tour is developed by connecting the person trips in a trip chain by time of day, travel activities and stop sequence. The figure below illustrates a hypothetical day pattern with two separate tours from / to home; and one sub-tour from / to work.
The model generates the travel for each individual in the study area, using micro-simulation. The demand model component is implemented in Python. The Model produces individual trip record output for each person in the model study area. These trip records are aggregated to give trip tables for use in EMME assignment. The model has 6 main components, applied to each person, as shown in Figure 2.
The model explicitly forecasts tours by tour purpose e.g. Work, School, Shop, Personal Business by each mode for different time periods in the day. It relies on the economic theory of individual choice behaviour and on the higher degree of precision available with a representation at the level of the individual traveller, meaning it can take into account (a) influences at the individual level, such as the walking distance to the bus stop for each person, (b) constraints at the household level, such as the number of cars in a household, time and money budgets, and (c) can consider specific activity patterns and therefore provide more accurate indications of behaviour.