Observer-Based Output Feedback MPC for T-S Fuzzy System With Data Loss and Bounded Disturbance

Area of Science:

• Control Engineering
• Systems Science
• Fuzzy Logic Systems

Background:

• Networked control systems (NCS) face challenges like data quantization and loss, impacting stability.
• Takagi-Sugeno fuzzy models are used to represent complex nonlinear systems.
• Bounded disturbances add further complexity to control system design.

Purpose of the Study:

• To develop an output feedback model predictive control (OFMPC) strategy for Takagi-Sugeno fuzzy NCS.
• To simultaneously address data quantization and data loss under bounded disturbances.
• To ensure closed-loop stability and satisfaction of input constraints.

Main Methods:

• Utilizing the sector bound approach to model quantization error as uncertainties.
• Modeling data loss as a time-homogeneous Markov chain.
• Applying S-procedure and quadratic boundedness for stability analysis and offline state observer design.
• Developing two online synthesis algorithms for OFMPC, including a novel formula for updating estimation error bounds.

Main Results:

• The proposed OFMPC strategy guarantees closed-loop stability for the fuzzy NCS.
• Input constraints are explicitly considered and satisfied.
• Recursive feasibility of the optimization problem is ensured through the updated ellipsoidal bound.
• Effectiveness demonstrated via a practical example.

Conclusions:

• The developed OFMPC techniques provide a robust solution for fuzzy NCS with simultaneous data loss and quantization.
• The novel algorithms enhance the reliability and performance of predictive control in uncertain networked environments.
• The methods offer a systematic approach to designing stable and feasible controllers for complex systems.