Performance evaluation of GPC algorithms based on different network-induced delay modeling methods

The networked control systems (NCSs) are now widely used in different industries with applications ranging from factory and home automation, to automotive and avionic control systems and to military and spatial applications. These NCSs have many attractive advantages (low cost, simple installation and maintenance, increased system agility, higher reliability and greater flexibility), but there is also a main disadvantage, the network-induced time-varying delay, that can deteriorate the closed-loop performances and even lead to instability. In this paper, firstly, the generalized predictive control (GPC) strategy is briefly presented and then the modeling methods for the network-induced time-varying delays are introduced. These methods assume that the delays induced by the communication network are time-varying, but bounded. Finally, the performances obtained by the networked predictive control strategy which is based on the different network-induced delay modeling methods are evaluated in order to illustrate its efficiency in controlling the clutch displacement of an electro-hydraulic actuated wet clutch.