| 摘要 |
An improved control for an automatic transmission upshift, wherein the on-coming clutch pressure (Ponc) is scheduled during the torque and inertia phases of the shift in accordance with the summation of feed-forward and feed-back control terms, the feed-forward control term being developed using an inverse dynamic model of the transmission. The feed-forward control term is determined by developing a desired acceleration trajectory of the input shaft, and periodically applying the acceleration trajectory to the inverse dynamic model to obtain an estimate of the required on-coming clutch pressure, given the transmission input torque. The feed-back control term is based on a comparison of the expected input speed response with the actual input speed response, and corrects for modeling errors in the feed-forward control, providing disturbance rejection and improved command following. Additionally, the closed-loop feedback error is used to enable adaptive correction of the feed-forward control so that the estimated on-coming clutch pressure more nearly produces the commanded input shaft acceleration. Controlling the input shaft acceleration in this manner achieves more consistent shift feel and energy dissipation in the on-coming clutch, with less intensive calibration effort and improved adaptability to different powertrain and vehicle-type configurations.
|
