Online identification and fusion estimation of road resistance coefficient based on FFRLS and EKF

To improve the accuracy of estimating the coefficient of road resistance for tracked vehicles on unpaved roads, this study proposes an online estimation method combining linear regression and nonlinear system analysis. The method utilizes Forgetting Factor Recursive Least Square (FFRLS) and Extended Kalman Filtering (EKF) to dynamically identify the coefficients for ground deformation resistance and slope resistance in real-time. Confidence levels are defined by calculating the standard deviations of both algorithms, and the estimation results from both methods are fused based on these confidence levels. The results demonstrate that the accuracy of this method in identifying road resistance coefficient (RRC) under flat and sloping conditions is 92.9% and 94.3%, respectively, significantly enhancing the precision of coefficient estimation.