Modelling inter-activity duration to capture weekly activity-travel dynamics: instilling inherent dynamics within daily travel demand models
本文提出一种Copula方法，联合建模个体一周内活动间持续时间和出行方式选择，利用加拿大GTHA地区一周出行日志估计模型，发现两者存在显著依赖关系，对理解多日出行行为有参考价值。
Explicitly modelling the multi-day dynamics can result in an accurate and unbiased understanding of activity-travel choices. The study introduces a copula approach to jointly model individuals’ inter-activity duration and travel mode choices using week-long travel diaries. This study uses the simulated likelihood technique to address the inherent left-censoring issue in inter-activity duration modelling using week-long travel diaries. The joint model is empirically estimated using week-long travel diaries collected in the Greater Toronto and Hamilton Area (GTHA), Canada. Seven mandatory and discretionary activity types are examined. The empirical model reveals a statistically significant dependency between inter-activity duration and travel mode choices. This dependency indicates that the frequent scheduling of specific activity types is positively correlated with higher utility from travel mode choices, particularly for retail shopping and service activities. Conversely, work/study and errand/grocery shopping activities show lower correlations, suggesting that higher mode choice utility has a lesser impact on the time between these activities.