Neural network estimation of soil reaction in rigid wheel and soft terrain interaction based on modified dynamic model

This study undergoes the application of neural networks in a modified wheel-solid interaction model. First, a new efficient model for interaction between wheel and soil is presented to overcome the deficiencies of traditional models for dynamic analysis. Moreover, as a novel technique, the neural network is briefly introduced and implemented in this analysis to estimate soil reactions. The results obtained in a simulation study reveal the potential of the presented approach for predicting soil reactions in various dynamic maneuvers.