Codesign of control and scheduling for networked control systems

Abstract

With the development of contorl theory and communicaton techonlogy, networked control systems(NCSs) have received more and more attention. The network that is introduced into the control systems produces some issues, such as network-induced delays, data packets dropout, network resources competition and so on. In this paper, we focus on the network-induced delays. However, the closed-loop control performance depends not only on the design of the control algorithm, but also on the appropriate scheduling of network resources. The codesign of single neuron PID-Smith predictive control and dynamic real-time EDF scheduling is proposed to solve the problems of NCSs. The single neuron PID-Smith predictive controller for NCSs has good learning ability to adapt the changes in the signal and time delays. The schduling theory is applied to allocate the limited resources to improve the utility of the network.