Pre-Shift fill control of wet clutches in heavy tracked vehicles using velocity prediction and a model predictive control framework

In heavy tracked vehicles equipped with wet clutches, clutch fill control during clutch-to-clutch shifting is crucial for enhancing shift quality and vehicle mobility. Existing studies have predominantly focused on clutch fill control after the shift command is issued. However, these post-shift control strategies fundamentally limit further performance improvements. Therefore, pre-shift clutch fill control becomes an important approach for further enhancing shift quality and vehicle mobility. This study introduces a pre-shift fill control strategy for wet clutches using velocity prediction and a model predictive control (MPC) framework. A comprehensive dynamic model of a heavy tracked vehicle is established, incorporating the power source, planetary-gear transmission, electro-hydraulic clutch actuation system, and load model. A velocity prediction algorithm is developed using historical on-vehicle data, integrating a Markov model, variational mode decomposition (VMD), and a radial basis function (RBF) neural network. Based on these predictions, a pre-shift fill control strategy based on the MPC framework is formulated to compute and update the target hydraulic pressure of clutches and brakes in real time through predictive optimization. Simulation and experimental results demonstrate that the proposed strategy effectively identifies shift timing and ensures reliable completion of the fill process. Compared with conventional post-shift clutch fill control, the proposed strategy improves shift quality, with the shift time reduced by approximately 17.8%.