Transit Origin Destination Estimation using Automated Data

Abstract

Development of an origin-destination (OD) demand matrix is crucial for transit planning. The development process is facilitated by transit automated data, making it possible to mine boarding and alighting patterns on an individual basis. This thesis presents novel methods for estimating transit OD matrix using automatically collected data. Depending on the type of transit automated data, there are two methods presented. A novel trip chaining method which uses Automatic Fare Collection (AFC), Automatic Vehicle Location (AVL), and General Transit Feed Specification (GTFS) data is proposed to infer the most likely trajectory of individual transit passenger. The method relaxes the assumptions on various parameters used in the existing trip chaining algorithms such as transfer walking distance threshold, buffer distance for selecting the boarding location, the time window for selecting the vehicle trip, etc. The thesis also proposes a method for estimating the transit route origin-destination (OD) matrix utilizing Automatic Passenger Count (APC) data. It uses $l_0$ norm regularizer, which leverages the sparsity present in the actual OD matrix. The technique is popularly known as compressed sensing (CS). The applications of both methods using automated data from Twin Cities, MN are also presented. The results show improved accuracy and more inference rate in calculating the OD matrix using trip chaining. Similarly, compressed sensing was found to work impressively well in evaluating transit route OD matrix within small errors.