System dynamic modelling and adaptive optimal control for automatic clutch engagement of vehicles

Abstract

A non-linear multi-rigid-body system dynamic modelling is developed for the automated clutch system in power transmission during clutch engagements, especially at sharp vehicle start-up, sudden engine flame-out, low driving speed and clutch plate overwearing. In order to guarantee an ideal dynamic performance of the clutch engagement, an adaptive optimal controller is designed by considering throttle angle, engine speed, gear ratio, vehicle acceleration and road condition. A non-linear model reference adaptive controller is utilized to prevent the engine from flame-out or fly-off effectively and the minimum value principle is also introduced to achieve an optimal dynamic performance of the non-linear system compromised in friction plate wear and vehicle drive quality. The optimal trajectory of the clutch engagement can be described in the form of explicit and analytical expressions and characterized by a deterministic and accurate control strategy instead of indeterministic and soft control techniques which need thousands of experiments. For validation of the controller, test work is carried out for the automated clutch engagements in a commercial car with a traditional mechanical transmission, a hydraulic actuator, a group of sensors and a portable computer system. It is shown through experiments that dynamic behaviours of the clutch engagement operated by the adaptive optimal control are more effective and efficient than those by a fuzzy PID control.