Feedback control through networks with packet loss: mixed H2/H∞ approach and application to a teleoperating system

We consider a networked control system (NCS) with packet loss being modeled as a stochastic process. We focus on the stochastic H∞ control approach. The designed controller is dynamic in the sense that it depends on the behavior of the lossy channel. First we apply the existing result on H∞ approach for NCS [4] to a teleoperating system. The fixed loss probability assumption in this result has shown a need of improvement through the simulation. Therefore we propose a stochastic H∞ control approach for a time-varying loss probability. The approach gives a dynamic controller which can be gain-scheduled according to the loss probability at each moment. Lastly, the idea of a stochastic H∞ approach is extended toward a mixed H2/H∞ version. Fig. 1: Bilateral teleoperating system