Indirect adaptive control of unknown multi variable nonlinear systems with parametric and dynamic uncertainties using...

Area of Science:

• Control Systems Engineering
• Artificial Intelligence
• Nonlinear Dynamics

Background:

• Adaptive control of unknown nonlinear dynamical systems presents significant challenges due to inherent uncertainties.
• Existing methods often struggle with systems having multiple inputs and states (MIMS).

Purpose of the Study:

• To develop an indirect adaptive regulation method for MIMS with dynamic and parameter uncertainties.
• To introduce a new neuro-fuzzy dynamical systems description for enhanced system modeling.

Main Methods:

• Utilizes a neuro-fuzzy dynamical systems description with fuzzy partitioning and high-order neural networks (HONN).
• Employs an indirect regulation strategy involving system identification around the operating point.
• Develops weight updating laws for HONNs to minimize identification error and system states.
• Constructs a control signal using a Moore-Penrose pseudoinverse for square and non-square systems.
• Introduces a parameter hopping method to ensure control signal existence, replacing conventional projection methods.

Main Results:

• Guarantees convergence of identification error and system states to zero.
• Ensures all closed-loop signals remain bounded despite uncertainties.
• The proposed control signal is applicable to both square and non-square MIMS.
• Demonstrates applicability through testing on well-known benchmarks.

Conclusions:

• The novel neuro-fuzzy approach effectively addresses indirect adaptive regulation for uncertain MIMS.
• The method provides robust control with guaranteed stability and bounded signals.
• The parameter hopping technique offers a reliable alternative for control signal existence in adaptive systems.