A theory is developed for the design of multiple-sensor weigh-in-motion (WIM) systems to minimise the errors caused by the dynamic axle loads of heavy vehicles moving at highway speeds. The theory is verified using measurements from a wheel load measuring mat of total length 38 m, incorporating 96 capacitive strip WIM sensors. The mat was installed on the Navistar test track in Indiana. A total of 460 test runs was performed on six different articulated heavy vehicles, at a range of speeds between 8 and 80 km/hr. The strip sensors were found to be reliable and to measure the dynamic wheel loads with errors of less than 4% RMS. The sensor calibration is independent of speed and temperature. The experimental results were found to agree closely with the theoretical predictions of multiple-sensor WIM performance and it was possible, using a three-sensor array, to measure static axle loads with approximately 6% RMS error, or less, for typical highway conditions of speed and road surface roughness. A good design for multiple-sensor WIM systems is to use three sensors, spaced evenly along the road. The sensors should be spaced according to a simple formula that depends only on the average traffic speed and the number of sensors.
Abstract