Optimal control input for discrete-time networked control systems with data dropout

These days, networked control systems (NCSs) in which data is transmitted via communication have been actively studied for many potential applications. In an NCS, data dropout degrades control performance depending on network conditions. For an NCS with data dropout, the authors propose a model-predictive-control-based input optimisation, representing data dropout as both a Bernoulli model and a finite-order Markov chain. Using the proposed NCS data dropout model, the authors derive an optimal input that provides the estimated error between the expected state of the plant and a given reference. The proposed control problem is formulated as its equivalent quadratic programming, as executed at each online sampling. The authors also demonstrate simulations and experiments to show the effectiveness of the proposed method.