ONLINE ESTIMATION OF WHEEL SINKAGE AND SLIPPAGE USING A TOF CAMERA ON LOOSE SOIL

摘要

A prominent concern with wheeled mobile robots that travel on loose and rough terrain is that they may get stuck in soil and experience a sharp increase in wheel sinkage and slippage. Various analyses of wheel-soil interaction mechanics have been conducted to study this issue. Several studies have been carried out for the online estimation of wheel sinkage and slippage using image processing. However, these image-based online estimation methods are greatly affected by target environment conditions such as lighting and reflectance. In contrast, a ranging sensor that provides three-dimensional point cloud data performs better at estimation even in an uncertain environment. This study proposes an online sinkage/slippage estimation method using a time-of-flight (ToF) camera that can measure the distance from a target object. Sinkage/slippage estimation of the traveling wheel on loose soil is experimentally demonstrated by attaching the ToF camera on the side of the wheel. The experimental result showed that most of the wheel sinkage could be estimated to within ±1 mm absolute error using point cloud data; traveling velocity and slip ratio of the wheel could be estimated for a resolution of less than 1 mm/s and less than 3.9% absolute error, respectively.