Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
The process results in an update of an agentâs needs and gaining knowledge from the experience. ⢠A prototype activity- based model of transport demand for Flanders, Belgium, called Feathers, will extend the Aurora model and add complementary con- cepts. The project is part of a wider research program involving a number of research institutes. Other elements being examined include the application of combined Global Positioning System (GPS) and personal digital assistant (PDA) technology for collecting activity- travel data (called PARROTS for PDA system for an Activity Registration and Recording of Travel Scheduling) and the use of new technology to collect vehicle data. ⢠Additional contributions to Feathers expected as part of the ongoing research program include modeling route choice behavior through the data obtained from PARROTS and calibrating the current model based on real- world data. Research will also test and improve currently used concepts of Aurora, such as estimating S- curves as utility functions, estimating the effect of con- text variables on maximum utility, evaluating the scheduling component, and extending learning facets. Additional concepts are also anticipated to be imple- mented in Feathers, including the impact of life trajec- tory events, which include events such as getting married and starting a job. Finally, research elements will focus on guiding and helping practitioners with the transition from four- step models to activity- based models. ⢠Aurora is an agent- based microsimulation system in which each individual in the population is represented as an agent. It is an activity- based model that simulates the full pattern of activity and travel episodes of each agent for each day of the simulated time period. The dynamics of the Aurora system start at the beginning of the day for each agent. The schedule is implemented based on the needs and knowledge of each agent. The environment has an impact on the implementation of the schedule for each agent in time and space. There is inter- action between agents who are competing against each other, which is when congestion occurs. Some agents decide to reschedule their original schedule based on their needs and knowledge. ⢠The scheduling and rescheduling model assumes that activities and travel are scheduled on a continuous time scale. The schedule meets a full set of scheduling constraints for each agent. Needs for activities grow over time and are satisfied by activities depending on dura- tion. Scheduling decisions are based on heuristics, rather than on an exhaustive search. Inputs to the scheduling model include utility functions, dynamic constraints, activity needs, and knowledge of the land use and trans- portation systems. ⢠The model is based on a set of utility functions. The utility of a schedule is defined as the sum of utilities across the sequence of travel and activity episodes that it contains. Utility is dependent on the time of the day, the activity duration, when the activity was performed, and the time since the previous activity. ⢠The input of the scheduling heuristic is a consistent schedule in terms of duration and timing choices. The output is also a consistent schedule with utility that is higher or equal to that of the inputs. The model itera- tively implements operations on an existing schedule until no further improvement is possible. Operations are evaluated under optimal duration and timing choices. Operations considered include inserting an activity, sub- stituting an activity, and repositioning an activity. Other possible operations are changing the location of an activ- ity, including or removing a return- home trip between activities, and changing the mode of a trip. ⢠Uncertainty is dependent on an agentâs attitude with respect to risk. Various decision- making principles can be accommodated within the model. Agents hold beliefs or subjective probabilities with respect to the expected state of system variables. Beliefs are represented by a probability distribution across possible system states. The expected utility of a schedule alternative is the weighted sum of the utilities of the schedule, depen- dent on the state variables, where the weights represent the beliefs. ⢠There are different types of learning. Attribute learning is the simplest form of learning. Agents learn about their environment based on their expectations. Agents update their beliefs about states of single system variables. Conditional learning relates to updating causal knowledge. For example, differences in travel time can be explained by the day of the week and the time of travel. Associative learning results from generalization. It means an agentâs beliefs can change or remain the same based on experience. Information- based learning is based on information sources such as the news media and agency announcements. The impact of this informa- tion depends on the credibility that agents place in the source. Social learning means that agents learn from members in their social network. ⢠Issues that have been identified to date relate to the synthetic population, belief updating, and other ele- ments. The system shows how an activity- based model can be used for microsimulation of spatial behavior. The framework embraces and integrates the urgency of activ- ities as a function of time, time budgets and competition between activities, spaceâtime constraints, the ability to reschedule activities, the ability to learn from interaction with the environment, and the ability to deal with uncer- tainty. The system allows users to analyze impacts of temporal as well as spatial variables on utilities and traf- fic flows. ⢠Other aspects will be added to the system within the context of Feathers. These aspects include explicitly 28 INNOVATIONS IN TRAVEL DEMAND MODELING, VOLUME 1
