Indirect Travel Time Data Validation Using Higher Order Crossings Analysis

Kaveh Farokhi Sadabadi,Ali Haghani,Lei Zhang
2012-01-01
Abstract:This paper considers the indirect validation problem of travel time data series using higher order crossings (HOC) analysis of time series. It is customary for traffic data validation purposes to directly compare “ground-truth” data obtained through representative field data collections with data under consideration to determine the quality and precision measures. This is a direct validation method which, due to the need for extensive field data collections, is prohibitively costly in large scale real-world applications. On the other hand, indirect methods either make use of readily available data such as existing archives (no field data collection) or perform sanity check to determine data validity. Therefore, indirect methods provide an economic alternative to efficiently assess the validity of large travel time datasets. HOC analysis approach is fully described and its application to indirect validation process is demonstrated using two alternative methods. Various real world test cases based on commercially available data are presented. Discussion as to advantages, limitations, and lessons learned from application of the proposed methods will be provided.
What problem does this paper attempt to address?