Algorithm Development for Derivation of Section-Related Measures of Traffic System Performance Using Inductive Loop Detectors

Abstract

Despite the advent of new detection systems such as video, infrared, microwave, and ultrasound, inductive loop detectors (ILD) still remain the most widely used sensors for traffic information. For example, the California Department of Transportation alone has approximately 300,000 ILD installations, and that number excludes the installations that are part of local cities and agencies. By using these ILDs in “smarter” ways, useful section-related traffic system parameters can be derived, including densities and travel times. Consequently, the existing detection infrastructure can be used for application in many areas of intelligent transportation systems and, especially, in real-time dynamic systems. The process of developing feature extraction and vehicle pattern matching algorithms and the subsequent derivation of section-related measures based on conventional inductive loops are discussed. Field data collected from State Route 24 in Lafayette, Calif., were used for algorithm development and for the testing of algorithm performance. The performance of a section-related algorithm is shown to be much better than the current method of extrapolation from local point measures.