Improving the Quality of Terrain Measurement

Abstract

The emergence of high-fidelity vehicle and tire models has raised the requirements for terrain measurement capabilities. Inaccuracies that were once tolerable for measurement of general terrain roughness are no longer acceptable for these new applications. The techniques in this work seek to improve the quality of terrain measurement in addition to providing an objective way to describe the accuracy of these measurements. The first portion of this work develops an accuracy verification procedure for terrain measurement systems. This procedure involves a static test to assess the limitations of the profiler's laser height sensor, and a corresponding dynamic test to evaluate the limitations of the positioning sensors. Even with a well calibrated system, inertial errors will accumulate. The second portion of this work develops techniques to address these inertial errors in the data post-processing phase. A general correction technique is developed for any terrain type and a more computationally efficient technique is developed for smooth surfaces.

For basic ride and handling simulations, 3D terrain surfaces are computationally impractical, as the models used for these simulations only require point excitation. Current road profilers acquiring these 2D data use single-point lasers that capture localized disturbances that would be mechanically filtered by the tire and suspension in the physical vehicle system. The final chapter in this work develops a method to extract a 2D terrain profile from a 3D terrain surface. By considering all of the information in the tire contact patch, the filtering properties of the tire are approximately emulated.