Byung Kwan Shin

Adaptive compensation control of vehicle automatic transmissions for smooth shift transients based on intelligent supervisor

In this paper, an advanced shift controller that supervises the shift transients with adaptive compensation is presented. Modern shift control systems for vehicle automatic transmission are designed to provide smooth transients for passengers’ comfort and better component durability. In the conventional methods, lots of testing and calibration works have been done to tune gains of the controller, but it does not assure optimum shift quality at all times owing to system variations often caused by uncertainties in shifting hydraulic systems and external disturbances. In the proposed control scheme, an adaptive compensation controller with intelligent supervisor is implemented to achieve improved shift quality over the system variations. The control input pattern which generates clutch pressure commands in hydraulic actuating systems, is updated through a learning process to adjust for each subsequent shift based on continuous monitoring of shifting performance and environmental changes. The proposed algorithm is implemented and evaluated on the experimental test setup. Results from the experimental studies for several operation modes show both improved performance and adaptability of the proposed shift controller to uncertain changes of the shifting environment in vehicle power transmission systems.

Estimation of Turbine Torque Using an Adaptive Nonlinear Observer for Closed-Loop Control of Vehicle Automatic Transmissions

This paper presents an adaptive sliding observer-based method for the estimation of turbine torque. The estimations of turbine torque can be used in the closed-loop control of vehicle power train for better shift quality. A sliding observer structure including saturation functions and boundary layers is applied to adapt torque converter parameters. The proposed method has been investigated via simulation and laboratory experimental studies. It has been shown via simulation and experiments that the proposed methodology is promising for the estimations of turbine torque since it uses only inexpensive angular velocity measurements and is not sensitive to parametric uncertainties.