On the implementation of a process control system on a CAN network: linking the process control parameters to the network parameters

Today communication networks are more and more used to implement distributed applications concerned by closed loop process controls. These applications require a timeliness service in order to guarantee performances (stability, ...). The goal of this paper is, by considering a closed loop process control application with an open loop transfer function of the type G(s)=K/s(1+taus) (both no position error and the control of the speed error), to discuss its implementation on the network CAN (which can be either dedicated or shared) i.e. to link, accounting for the desired performances for the control application (phase margin, overtaking, rise time), the parameters associated to the process control (K, tau) and to the network (frame duration, priorities, buffering)