A large-scale survey for household weekend activity and related travel was completed recently in the City of Calgary. The data include detailed information of traveler and activity, such as personal type (e.g., adult worker or senior), employment status (fulltime or part-time), annual income, gender, activity type (e.g., shopping or sociality), activity duration, and starting & ending time of each activity. A micro-simulation based choice behavior model has been used in the previous city planning tasks. The model is capable of simulating complete travel behavior of individuals by considering travel purpose, travel mode, itinerary, activity durations, and even group influences. Previously, the simulation was done using a Monte Carlo process with sampling distributions based on weighted sample of observed durations. Simulations based on such “static” distributions, however, can not be used to analyze the influences of various policies (e.g., changes in transit fare) and travel conditions (congestion or easier accessibility) to household activities in a dynamic environment. This study is an initiative for modeling the relationship between activity durations and various influencing factors (e.g., personal type, employment status, and income level, etc.). Especially, hazard and survival functions are specified for each type of activity and individual personal type. The results show high degree of fit. It is believed that these models would be useful for travel-related policy analysis in the future modeling framework.
