Evaluation of Speed Measurement and Prediction Techniques for Signalized Arterials

Abstract

The ability to accurately measure and predict the average speed of traffic for different facilities and conditions has become increasingly important with the implementation of the Environmental Protection Agency's Transportation Conformity Rule. Average speed is a vital input into motor vehicle emission rate models. Three speed measurement techniques (floating cars, loop detectors, and aerial photographs) and two average travel speed prediction techniques (the Highway Capacity Manual method for arterials and TRANSYT-7F for signalized arterials) are evaluated. These techniques are compared and evaluated against 4 hr of field data gathered for an 8-mi (13-km) section of Ventura Boulevard in Los Angeles, California. The mean speeds measured by floating cars and aerial photographs were in close agreement (within 6 percent of each other). Loop detectors generally produced mean speeds 22 percent higher than the floating car measurements, even after correcting for the differences between the space mean speeds measured by floating cars and the time mean speeds measured by the loops. TRANSYT-7F was able to predict the average travel speeds on the arterial within 3 percent of the floating car measurements and 5 percent of the aerial photograph measurements. The Highway Capacity Manual estimates were uniformly 18 percent lower than the measured floating car speeds. A correction to the method improved the estimates but tended to slightly over-compensate. The adjusted estimated speeds were 9 percent higher than the floating car speeds.