The Simulation of Activity Diary Data Using Sequential Probability Distributions

This paper has developed and evaluated a framework for simulating activity diary data. It is claimed in the paper that simulating activity diary data can be useful to assist practitioners and researchers in the estimation of activity-based travel demand models. Especially in cases were the cost of collecting activity diary data is huge, or in cases when response rates are low, the simulation and creation of synthetic diary data is believed to be a viable alternative or supplement to collecting data. In this paper, a new methodology to simulate activity diaries was therefore developed and empirically tested. The presented approach shows several ways to store the sequential information (sequences of activities and travel) which is typically incorporated in an activity diary. The approach is novel, especially with respect to store information in “codebooks”, a term which is introduced to reflect that the information which is kept, represents the combinations of activities that typically sequentially occur in a persons’ diary. Based on this sequential information, sequential probability distributions were constructed in order to simulate new and real-world activity diaries. The diaries were generated by means of Monte Carlo simulation and the empirically derived sequential probability distributions were used as a constraint in the simulations. Sequential information which goes back in time up to several reported activities, is stored. In order to make a mature evaluation of the simulated diaries, different performance indicators were considered by using detailed pattern-, tripand activity-level analyses. It is shown in the paper that the results are satisfactory and that the initial framework developed in this paper holds out considerable promise for simulating activity diary data.