State estimation for distributed drive agricultural electric vehicles in complex mountain environments

Abstract

Accurate and reliable state estimation is a fundamental prerequisite for ensuring the high-precision autonomous navigation and driving stability of large agricultural machinery during complex field operations. To address the sharp accuracy degradation of traditional algorithms under severe nonlinear dynamics and abnormal measurement disturbances, an enhanced square-root cubature Kalman filter (SRCKF) is proposed. Building upon a three-degree-of-freedom (3-DOF) vehicle dynamics model, the proposed algorithm integrates physically guided state constraints, a conservative fixed gain attenuation strategy, and a dual-layer abnormal-measurement protection mechanism. Consequently, integration drift and numerical instability under severe road disturbances and abnormal measurements are effectively suppressed. Through CarSim–Simulink joint simulations, it is demonstrated that the proposed method achieves accurate tracking of vehicle motion states together with improved disturbance rejection under adverse conditions such as chassis twist and discrete impacts. Quantitative analysis reveals that the proposed algorithm suppresses the root mean square error (RMSE) of critical states, such as lateral velocity, to a magnitude of 10 −2  m/s under extreme off-road excitations, showing lower estimation errors than the standard SRCKF baseline under the investigated simulation conditions. The results demonstrate the potential of the proposed estimator for vehicle-state estimation under the investigated mountainous off-road simulation conditions.