Dynamic hybrid model of an electro-pneumatic clutch system

• Published in: Mechatronics (Oxford) Vol. 23; no. 1; pp. 21 - 36
• Main Authors: Szimandl, Barna, Németh, Huba
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01.02.2013
• Subjects: Clutches; Design engineering; Dynamic hybrid model; Dynamical systems; Dynamics; Electro-pneumatic clutch; Mathematical analysis; Mathematical models; Modelling; Solenoid magnet valve; State space realisation; Validation; Verification; Validation; Verification; Dynamic hybrid model; Solenoid magnet valve; Electro-pneumatic clutch; State space realisation
• ISSN: 0957-4158; 1873-4006
• DOI: 10.1016/j.mechatronics.2012.10.006

► We model an electro-pneumatic clutch system, which is used for medium- and heavy duty commercial vehicles. ► The model is constructed on the basis of the conservation, i.e. first engineering principles. ► It is transformed into state space form for dynamic simulation, parameter estimation and control design/validation purposes. ► The verification and validation of the developed model have been carried out. This paper deals with modelling an electro-pneumatic clutch system, which is used for medium- and heavy duty commercial vehicles. The mathematical model is built up for dynamic simulation, parameter estimation and control design/validation purposes, which is a phase of the design process of a new clutch system. These intended applications define the modelling goals and determine the modelling assumptions, which let one to reduce the model complexity. Since the model shows discrete–continuous behaviour, i.e. the model has hybrid properties, a nominal state domain or hybrid mode has been chosen for the sake of simplicity, where the model is continuous. In addition all the cases are given systematically, where the model has discrete transients. The model is constructed on the basis of the conservation principles such as mass, energy, momentum and magnetic linkage conservation and it is provided with constitutive equations to get a solvable set of equations. This final collection is then transformed into state space form for the given applications above. The verification of the developed model is carried out using extensive simulations against engineering perception and operation experience on the qualitative behaviour. Then for validation purposes the outputs of the model are compared to measurements on the real system to give a quantitative performance index about the model accuracy. Since for model-based controller design the developed model is too complex it should be simplified. Hence possible model reduction methods are proposed, which omit all details that are weakly represented in the state variables/outputs and not coupled with the control aims.