Coordinated dual-rolling control of shifting in HMCVT based on KF-MPC algorithm with anti-disturbance capabilities

•Proportional hydro-mechanical transmission is suited to wheeled tractor features.•The Kalman filter improves the accuracy of disturbed variables in real time.•The shifting process disturbances can be quantified as control variables.•Kalman filter real-time correction of input variables for model predictive control. The hydraulic-mechanical continuously variable transmission (HMCVT) integrates hydraulic step-less speed regulation and mechanical efficient transmission characteristics, a key component to ensure the efficient and stable operation of agricultural machinery such as tractors. As intelligent agriculture advances, the demand for smooth and continuously variable speed capabilities in HMCVT systems is steadily rising. To address the issues of large shift impact and poor stability caused by shift disturbances during the variable speed process of HMCVT, a dual rolling coordination control strategy, which integrates Kalman Filter (KF) and Model Predictive Control (MPC), is proposed. This inquiry focuses on the shifting process of the HM1-HM2 segment within the proposed proportional two-stage hydro-mechanical continuously variable transmission (2S-PHMCVT) structure. Firstly, we establish dynamic models of the shift process in each transmission mechanism. The shift disturbances are then analyzed based on transmission principles and quantified into controllable variables. Secondly, a control principle model for shift-related variables is constructed, and a KF-MPC joint controller is designed based on the relationship between control variable output characteristics. Finally, we establish an AMEsim-Simulink joint simulation platform for validating the filtration effectiveness of the KF and the shifting quality. Additionally, a shift test platform is set up for further test validation. The results indicate that the KF estimates values close to the actual ones, demonstrating the mean square error of only 0.1223. Compared to MPC control and no control, adopting KF-MPC control reduces the maximum impact of the system’s shift process by 48.5 % and 66.9 %, respectively, while reducing friction work by 23.4 % and 42.2 %, and advancing the shifting time by 13.1 %. This study utilizes the filtering properties of the KF to mitigate shift disturbances in HMCVT, thereby dynamically correcting the input variable information of MPC in real time, ultimately achieving the joint enhancement of shift quality. The research findings can provide a new theoretical foundation and