Abstract Manager

Travel demand models express complex interdependencies between sequences, choices, market segments and/or decision-makers. These models have been traditionally implemented in a range of different technologies, tools, and languages. Even within a single agency, modelers may be required to master knowledge of different programming languages and operating conventions simply to work across model applications. Without any real separation of model configuration and software implementation, too much complexity is exposed to the users resulting in longer model development cycles. Moreover, irrespective of the travel demand model structure, these demand models undergo a lengthy calibration process which requires calibration across several data sources. The calibration process has been ad-hoc and manual without any systematic diagnosis of the data used for the calibration.

The City of Ottawa is upgrading the TRANS Regional Transportation model using AGENT demand modeling platform. AGENT separates the model structure from the model platform using a concept of a model package. This acts as a general specification of travel demand modelling and enables modelers to formally describe virtually any travel demand model, completely independent of any implementation details. The separation of model structure from the model platform significantly reduced the model configuration period from months to a few weeks. This presentation includes the details of the workflow used to configure the new TRANS model.

AGENT also makes it possible for modelers to express the rules required for a new automated calibration method that works with virtually any travel demand model structure, from four-step to ABM. Instead of adjusting choice models sequentially, as is done in current and previous practice, automated calibration operates to adjust model coefficients, as configured by modelers, simultaneously on the entire model system for quicker, more consistent, and finer calibrations. Instead of being limited to only one data type, automated calibration can fuse multiple, disparate data sets including household travel surveys, “big data” O-D and traffic counts, transit fare card data, and more by leveraging the travel model as a common denominator. The new TRANS model was calibrated simultaneously against highly segmented behavioral targets from the 2022 O-D survey, time-of-day traffic counts, and time-of-day transit boarding counts in a short time frame. The full presentation will include key automated calibration configurations and the calibration results.