Modelling Stability in Travel Behaviour

交通需求模型是交通规划中对交通设施或交通政策进行评估不可或缺的工具。常见的出行需求模型的分析周期一般为一天。然而，旅行行为研究发现，为了观察旅行的某些特征，需要进行为期数天的调查。一天的观察期通常太短，无法观察旅行行为的变异性和稳定性。在模态选择的情况下，稳定性可以用单模态旅行行为来识别，而可变性可以用多模态旅行行为来识别。对于运输规划来说，这种见解很重要，因为可以假设，单式运输的人（模式选择的高度稳定性）与多式运输的人（模式选择的高度可变性）对旅行供应变化的反应不同。例如，一个经常使用汽车的人可能比偶尔使用公共交通工具的人受到公共交通改善的影响更小。该项目旨在通过纳入目的地选择和模式选择行为的稳定性来提高出行需求模型的现实性。为此，我们将研究几种在目的地选择和模式选择中模型稳定性的方法。所得模型将在微观旅行需求模拟中进行测试。根据描述旅行行为稳定性的指标，不同模型的结果将相互比较，并与德国流动性小组和区域家庭旅行调查的数据进行比较。基于这些指标的比较可以评估各个模型在再现旅行行为稳定性方面的表现。将在本项目范围内确定和制定描述旅行行为稳定性的适当指标。该项目的结果将是洞察如何通过纳入目的地选择和模式选择行为的稳定性来改善旅行需求模型的现实性。可以预期，旅行行为稳定性的改进再现也提高了这些模型所作预测的质量。

Travel demand models are an indispensable tool for the assessment of transport facilities or policies in transport planning. The analysis period of common travel demand models is typically one day. Travel behavior research, however, has found that multi-day surveys are necessary to observe certain characteristics of travel. An observation period of one day is typically too short to observe variability and stability in travel behavior. In the case of mode choice, stability can be identified with a monomodal travel behavior, while variability can be identified with a multimodal travel behavior. For transport planning this insight is important since it can be assumed that a monomodal person (high stability in mode choice) reacts differently than a multimodal person (high variability in mode choice) to a change in travel supply. For example a person who always uses the car will probably be less affected by an improvement in public transport than a person who occasionally uses public transport. The project aims at improving the realism of travel demand models by incorporation stability in destination choice and mode choice behavior. For this purpose several approaches to model stability in destination choice and mode choice will be examined. The resulting models will be tested in a microscopic travel demand simulation. Based on indicators describing stability in travel behavior, the results of the different models will be compared with each other and with the data of the German Mobility Panel and a regional household travel survey. The comparison based on these indicators allows assessing the performance of the individual models in regard to reproducing stability in travel behavior. Suitable indicators to describe stability in travel behavior will be identified and developed within the scope of this project. The result of the project will be the insight how realism of travel demand models can be improved by incorporating stability in the destination choice and mode choice behavior. It can be expected that the improved reproduction of stability in travel behavior also improves the quality of the forecasts made by these models.