T-S fuzzy-model-based output feedback MPC for networked control systems with disturbance

This paper investigates output feedback model predictive control (MPC) for nonlinear networked control systems (NCSs) with bounded disturbance where data quantization and packet loss may occur simultaneously. A Takagi-Sugeno (T-S) fuzzy model is exploited to describe the nonlinear system which can be turned into a linear one. The quantization error is treated as sector bound uncertainties and a binary Markov chain is introduced to characterize packet loss of NCSs. By applying the notion of quadratic boundedness (QB), a state estimator can be off-line designed and the estimation error bound can be obtained. The on-line MPC optimization problem which minimizes an upper bound of the expect value of the infinite horizon performance cost is solved based on the obtained estimation state. A new technique for refreshing the estimation error bound, which plays the key role of guaranteeing the recursive feasibility of optimization problem, is provided. A numerical example is given to illustrate the effectiveness of the proposed output feedback MPC approach.