Modeling Urban Residents' Activity Chain Choices And Travel Behavior

Using the activity-based modelling approach, this thesis studies the impacts of fuel surcharge policy and online shopping on residents' daily activities and travel behaviors, and then studies the existence of paradox. This study can provide fundamentals for estimating the effects of fuel surcharge policy and online shopping on the transportation network as well as households'behaviors, and can also serve as a useful tool for avoiding inappropriate investments in transportation planning.Firstly, this thesis proposes an analytical model to determine the optimal combination of fuel surcharge and transit frequency so as to maximize the social welfare of a time-varying traffic network with multiple travel modes. In the proposed model, the effects of fuel surcharge and transit frequency on commuters'travel costs are explicitly considered. Each user aims to determine the combination of the activity chain, activity start time, trip departure time, travel path, and transportation mode so as to maximize its own net benefit, which is defined as the gap between the utility gained from the activity and the cost during travel. A heuristic algorithm is then developed to determine the optimal fuel surcharges and bus frequency simultaneously. In addition, the evaluations of the impacts of the fuel surcharge policy on residents' activities, the system energy consumption, and social welfare are carried out by comparing the system performances before and after implementation of the fuel surcharge policy.Secondly, this thesis presents a theoretical model to analyze the effects of online shopping on the individual activity trip chain. In the proposed model, two types of retailers are considered:online and brick-and-mortar retailers. Individuals first choose their preferred shopping channels (online shopping or traditional shopping) based on the utility level, and then determine the activity trip chain, activity duration and total goods consumption. Before the emergence of e-commerce, all users have to go to brick-and-mortar stores to purchase products to maximize their own utility, and the user equilibrium will be achieved. After the emergence of e-commerce, residents determine the shopping channel to maximize their own benefit, and the new user equilibrium will be achieved. The user equilibriums before and after the emergence of e-commerce are compared, in terms of the total traffic volume, residents'travel behaviors, activity chain, and activity duration. Also, the effects of the government tax policies on households' activities, the revenues of physical and online retailers, and the total traffic volume are analyzed.Finally, the thesis investigates the Braess-like paradox phenomenon in a transit network using the activity-based modelling approach. Both the activity-based model and trip-based model are developed to simulate users'behaviors. A time-dependent activity and travel choice equilibrium model is proposed, which simultaneously considers a traveler's choices of trip chain, departure time, route, and activity timing and duration over the time of day. The model is then formulated as an equivalent variational inequality problem. Numerical results show that the activity-based model and trip-based model react differently to the change of the transit frequency and marginal activity utility in terms of the existence of paradox. In addition, when using the activity-based model, an increase in the marginal activity utility would lead to a decrease in the region of the transit frequency settings within which paradox occurs. Furthermore, the paradox region is also provided in this thesis, which is jointly determined by the marginal activity utility in demand side and the transit frequency in supply side.