Predicting Maximum Traction to Improve Maneuverability for Autonomous Mobile Robots on Rough Terrain

This paper proposes a method to predict maximum traction force for autonomous mobile robots on rough terrain in order to improve maneuverability. For predicting traction force, we utilize friction-slip curves based on modified Brixius model derived empirically in terramechanics, which is a function of mobility number Bn′ and slip ratio S. Friction-slip curves include physical characteristics of various rough terrains such as firm soil, sandy soil and grass-covered soil. Also, we build prediction models for terrain parameters; maximum static friction and optimal slip ratio on friction-slip curves. Mobility number Bn′ is estimated from modified Willoughby Sinkage model which is a function of sinkage z and slip ratio S. Therefore, if sinkage z and slip ratio are measured once by sensors such as a laser sensor and a velocity sensor, then mobility number Bn′ is estimated and maximum traction force is predicted from a prediction model for terrain parameters. Estimation results for terrain parameters are shown in a driving simulation using MATLAB. Prediction performances for maximum traction of various terrains are evaluated as high accuracy through analysis of estimation errors. 