Network based controller applied to a highly dynamic system

The paper deals with the design of the real-time net-worked control system (NCS) for the experimental magnetic levitation system (MLS) which is an example of the highly dynamic and structurally unstable controlled plant. The experimental system under consideration consists of the plant (MLS) equipped with sensor, actuator and a local computer responsible mainly for communication with a remote controller computing values of control feedback and connected with the local computer by a computer network. The design of this NCS is performed using the predictive control approach which is relatively simple for implementation and well suited for systems with variable transmission delays, even several times longer then the sampling period used for discretization of the plant model. In this paper we assume that, the network is a general purpose network without fixed or even constrained latencies.