Estimation of vehicle clutch torque using combined sliding mode observers and unknown input observers

Precise clutch control is a crucial factor in determining the ride quality of the production vehicle and its performance has been enhanced by accurate estimation of involved clutch torque or pressure, especially during gear changing phase (torque phase / inertia phase). Unfortunately, torque or pressure sensor is not available in production cars because of its cost, installation, and maintenance issues. This study mainly focuses on the accurate estimation of the clutch torques for the vehicles with automatic transmission using sliding mode observers and unknown input observers. Sliding mode observer is selected because of its applicability to nonlinear systems and characteristics of its robustness to model uncertainties. The required values for estimation include the engine map, torque converter characteristic curve, and various vehicle parameters such as gear ratio and inertia. Using the above-mentioned information, sliding mode observer is constructed without additional sensors. The proposed observer is validated by computer simulation using commercial software Matlab & Simulink. The estimated output shaft torque can be used as a reference for optimal clutch control and torque sensor which is not possible in the production car can be replaced with those developed observers.