Genetic-Algorithm-Assisted Sliding-Mode Control for Networked State-Saturated Systems Over Hidden Markov Fading...

Area of Science:

• Control Systems Engineering
• Stochastic Systems
• Communication Networks

Background:

• State-saturated systems present control challenges, especially over unreliable communication channels.
• Existing fading channel models lack the generality and practicality for complex network scenarios.
• Accurate estimation of network conditions is crucial for robust control system design.

Purpose of the Study:

• To develop a sliding-mode control (SMC) strategy for state-saturated systems operating over time-varying fading channels.
• To introduce a novel hidden Markov fading channel (HMFC) model for enhanced practical applicability.
• To ensure system stability and reachability within a defined sliding region despite channel uncertainties.

Main Methods:

• Modeling fading channels as a finite-state Markov process.
• Employing a hidden Markov mode detector to estimate the current network state.
• Designing a switching-type SMC law based on estimated network modes.
• Utilizing stochastic Lyapunov stability and hidden Markov models to derive stability conditions.
• Developing a binary genetic algorithm (GA) for controller design, optimizing for reachability.

Main Results:

• The proposed HMFC model is more general and practical than existing models.
• Sufficient conditions for mean-square stability and reachability are established.
• A GA-assisted SMC scheme is successfully developed and verified through a numerical example.
• The controller effectively handles state saturations and channel fading.

Conclusions:

• The developed GA-assisted SMC scheme provides a robust solution for state-saturated systems over HMFCs.
• The novel HMFC model and control strategy offer improved performance and applicability in realistic communication environments.
• The study demonstrates the effectiveness of integrating hidden Markov models and genetic algorithms in advanced control design.