Online adaptive identification of clutch torque transmissibility for the drivability consistency of high-performance production vehicles

Abstract

Drivability in vehicles with a wet-type dual-clutch transmission (DCT) relies heavily on the precise torque transfer control of the clutch to efficiently distribute the engine torque to the wheels. However, the large variation in the clutch torque transmissibility caused by temperature and wear and the lack of torque sensors for the driveline in production vehicles are the two significant challenges that hinder accurate clutch control in wet DCT vehicles. Therefore, knowledge regarding clutch torque transmissibility is crucial for realizing the demanded clutch torque because it can help determine the required clutch pressure. Thus, in this study, an online adaptive identification algorithm for clutch torque characteristics is developed for high-performance vehicles with wet-type DCTs. The proposed algorithm estimates each clutch torque characteristic using the extended Kalman filter (EKF) based on a parameterizable model for the torque transmissibility curve, making it applicable to production vehicles. Further, the algorithm incorporates an adaptation law based on the Lyapunov stability theorem to compensate for uncertainties in the engine torque, thereby ensuring a robust identification performance regardless of the installed engine type. The estimation performance and effectiveness of the adaptive identification algorithm are experimentally validated using test vehicles with wet DCT.