Fault-tolerant control design for unreliable networked control systems via constrained model predictive control.

Area of Science:

• Control Systems Engineering
• Networked Systems
• Stochastic Systems

Background:

• Networked control systems (NCSs) are susceptible to imperfections like random time delays and packet dropouts.
• These network issues can be modeled using Markov chains, leading to Markovian jump linear systems (MJLS).
• Uncertainties in the transition probability matrix (TPM) and practical fault models add complexity to NCS control.

Purpose of the Study:

• To develop a passive fault-tolerant control (FTC) strategy for discrete-time NCSs.
• To address network imperfections (time delay, packet dropout) and unknown TPM elements.
• To incorporate a comprehensive fault model and input constraints into the control design.

Main Methods:

• Modeling network imperfections as a Markov chain for MJLS.
• Employing a state augmentation technique to obtain the closed-loop NCS model.
• Proposing a constrained model predictive control (MPC) strategy.
• Deriving sufficient design conditions using linear matrix inequalities (LMIs).

Main Results:

• A reliable fault-tolerant controller was designed considering network uncertainties and faults.
• The proposed MPC strategy effectively handled input constraints.
• Simulation examples validated the controller's effectiveness and superior performance compared to existing methods.

Conclusions:

• The developed passive FTC strategy is effective for discrete-time NCSs with network imperfections and faults.
• The use of MPC and LMI-based conditions provides a robust control solution.
• The proposed method offers improved performance over current state-of-the-art approaches.