Identification of soil parameters for unmanned ground vehicles track-terrain interaction dynamics

Abstract

This paper presents a novel method for identifying soil parameters during traversing of a tracked vehicle on unknown terrain. Two identification methods, Newton Raphson (NR) method and least square (LS) method are used and compared for prediction accuracy and computational speed. The NR method identifies unknown individual parameters of the soil and can be used when the number of equations is the same as the number of parameters to be identified. The LS method optimises the root-mean-square residual error of the model. The methods are used to identify the soil parameters, cohesion and internal friction angle. These identified soil parameters can be used to increase the autonomy of a tracked vehicle. Results show that NR method is better in terms of prediction accuracy and computational speed. The NR method, therefore, is more suitable for online identification of soil parameters. This paper shows that, in case of heavy tracked vehicle, soil cohesion has small effect on the vehicle performance. This is advantageous in term of increasing computational speed.