A Novel Method of Soil Parameter Identification and Force Prediction for Automatic Excavation

Abstract

A shortage of soil property parameters is preventing the automatic excavation system from being used in a digging operation adaptively like an operator. This study proposes a new method of soil parameter identification and digging resistance prediction that classifies the entire excavation process into three parts: penetration, cutting, and loading. A fuzzy estimation strategy is introduced into penetration to estimate the property parameters of the soil depending on the peak value and the average value of soil resistance. Furthermore, an improved FEE model was used to describe the soil-tool interaction and predict the soil resistive forces. The traditional way of predicting after identifying, which needs at least two excavations, can hardly equal this method in flexibility and real-time functionality. The proposed method can identify soil parameters and predict digging forces on one excavation, which is more adapt to soil parameter variation. The experimental results show that the proposed method can identify precisely and predict effectively.