Comparison Between Two Estimation Models for the Friction Torque of Dry Clutch

The friction torque of dry clutch is estimated under a condition without extra added sensors based on easily acquired signals of conventional mass-production vehicles, i.e. the rotational speed and nominal torque of engine and the nominal pressing force of clutch. Both gradient projection algorithm-based and adaptive estimation models are established, whose performances are then analyzed by simulation and bench test. The results show that adaptive estimation model exhibits higher accuracy in estimating the friction torque of clutch.