Adaptive Model Prediction of Unmanned Agricultural Machinery for Tracking Control in Mountain Environment

Abstract

In order to improve the trajectory tracking accuracy and body stability of unmanned agricultural machinery in mountainous environment, this paper designs the adaptive forgetting factor related to the driving state of the agricultural machinery, and then corrects the tire turning stiffness in real time based on the Adaptive Forgetting Factor Recursive Least Squares (AFFRLS) algorithm. and adaptively adjust the weight coefficients in the MPC according to the road surface attachment coefficient to achieve the dynamic control between the trajectory tracking accuracy and the lateral stability of the body of the unmanned agricultural machinery in the mountainous environment. The results show that the trajectory tracking accuracy and lateral stability of unmanned agricultural machinery in mountainous environment can be dynamically controlled compared with the previous methods. The results show that the proposed adaptive variable-parameter MPC algorithm (AMPC) control algorithm improves the tracking accuracy and stability compared to previous trajectory tracking control algorithms, resulting in a reduction of 36.1% and 26% in the peak beta and yaw rate of the unmanned agricultural machinery, respectively, and a reduction of 67% in the peak lateral tracking error.

Keywords

• Unmanned agricultural machinery
• tire cornering stiffness estimation
• AFFRLS
• tire slip angle constraints
• adaptive weight coefficient