Modeling Visit Probabilities within Space-Time Prisms of Daily Activity-Travel Patterns

Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

Abstract

As a central time geographic concept, space-time prism (STP) delimits the space-time opportunities that can be reached by a moving object and is widely applied to measure the ability of individuals to travel and participate in activities. The seminal STPs are binary measures such that all locations are considered equally accessible if within the prisms, otherwise not accessible. In reality, a prism does not have homogeneous interiors. The notion of visit probability was raised to describe the likelihood of visiting different locations within the prism, in which locations with small visit probabilities are unlikely to be accessed. However, the previous studies modeled probabilistic STP at the trip level for conducting single activities. Such probabilistic characteristics of accessibility within a prism have not been investigated for daily activity programs (APs) that potentially include multiple activities with flexible activity sequences. This study proposes an approach of constructing and estimating the probabilistic STP of an AP. First, we construct the activity-based STP for a daily AP based on the multi-state supernetwork. Second, we model the visit probability within the activity-based STP using semi-Markov techniques. Third, we estimate the visit probabilities based on GPS trajectories and simulate the empirical visit probability to demonstrate the validity and effectiveness of the approach. The estimated visit probabilities at the locations can be applied for dynamic choice set generation of ATPs and more realistic accessibility analysis.
Original languageEnglish
Title of host publication4th International Time Geography Conference
Publication statusAccepted/In press - 1 Jun 2022

Fingerprint

Dive into the research topics of 'Modeling Visit Probabilities within Space-Time Prisms of Daily Activity-Travel Patterns'. Together they form a unique fingerprint.

Cite this