The detection of long-distance pavement elevation undulation is the main data basis for pavement slope detection and flatness detection, and is also the data source for 3D modeling and quality evaluation of pavement surfaces. The traditional detection method is to use a level and manual coordination to measure; however, the detection accuracy is low and the detection speed is slow. In this paper, the high-speed non-contact vehicle-mounted road undulationelevation detection method is adopted, combined with the advantages of each sensor measurement; three methods are proposed to detect the road undulation elevation: rotary encoders, accelerometers, attitude sensor data fusion detection; GPS RTK detection; and Kalman filtering detection. Through modeling and experimental comparison, Kalman filter detection is not disturbed by the environment, and the detection accuracy is higher than the current international standard.
