Abstract
Most existing traffic flow models rely on data collection methods that require a detailed layout of networks with compilations of recorded individual trip data. Although these procedures have been reliable, they also possess disadvantages such as high computation costs and a lack of privacy protection. Thus, in search of a lower cost alternative that can also effectively protect consumer privacy, we analyzed the Bathtub traffic flow model as a potentially viable data collection protocol.To test whether concepts can be proven, conservation equations can be consistent, and outputs can be obtained with accuracy through the Bathtub model, I performed model calibration and validation on data provided by Bluebikes, Metro Boston’s public bike share program. The following components were tested: unified relative space paradigm, conservation equations, and Bathtub model. These components were tested through the following steps: data organization, definition of steps, Bathtub model selection, Bathtub variables, Bathtub relative variables, average speed, conservation equation validation, and model solution. The unified relative space paradigm unified the network trips using remaining trip distances. Bluebikes trip distance distribution showed a log-normal distribution, which failed to meet the negative exponential and time-independent trip distance distribution assumption. The conservation in total trips equation was validated with perfect accuracy, while the conservation in trip-miles-traveled equation was validated with good accuracy. The generalized Bathtub model solution also produced accurate results, where space-mean speed yielded the best results. Given the model’s novelty and potential for privacy-preservation and application, there are many possibilities for future study, such as: data collection protocols with the Bathtub model, compatibility with other transportation modes, and comparisons with reality. This study establishes the preliminary step in putting theory to practice, as we aim towards application.