Neural-Networks-Based Prescribed Tracking for Nonaffine Switched Nonlinear Time-Delay Systems

Area of Science:

• Control Engineering
• Nonlinear Systems
• Artificial Intelligence

Background:

• Traditional adaptive control struggles with unknown nonaffine switched nonlinear time-delay systems due to indifferentiable terms and relaxed controllability conditions.
• Existing methods often fail to efficiently address the complexities of these systems, limiting tracking performance.
• The need for robust control strategies for systems with time delays and nonaffine dynamics is critical in many engineering applications.

Purpose of the Study:

• To develop an adaptive control scheme for unknown nonaffine switched nonlinear time-delay systems.
• To overcome the limitations of traditional adaptive control algorithms in handling indifferentiable nonaffine terms.
• To achieve prescribed tracking performance while ensuring system stability.

Main Methods:

• Utilizing neural-networks (NNs) for separation and approximation of nonaffine functions.
• Employing dynamic surface control and convex combination techniques for controller and switching strategy design.
• Incorporating an adaptive law for each subsystem to minimize conservativeness.

Main Results:

• Successfully designed an adaptive controller and switching strategy for the specified class of systems.
• Demonstrated that all signals in the closed-loop system remain bounded.
• Achieved the prescribed tracking performance level for the nonlinear time-delay systems.

Conclusions:

• The proposed neural-network-based adaptive control scheme effectively addresses unknown nonaffine switched nonlinear time-delay systems.
• The method overcomes the challenges posed by indifferentiable terms and relaxed controllability conditions.
• The developed approach guarantees system boundedness and achieves desired tracking performance, offering a significant advancement in control theory.