Low peaking-phenomenon cascade high-gain observer design with LPV/LMI method.

Area of Science:

• Control Systems Engineering
• Nonlinear Systems Analysis
• Observer Design

Background:

• High-Gain observers are prone to peaking phenomena and noise sensitivity.
• Existing methods often struggle to mitigate these inherent limitations effectively.
• Accurate state estimation is crucial for control and monitoring applications.

Purpose of the Study:

• To propose a novel parameter tuning method for a 2nd-order cascade observer structure.
• To address the peaking phenomenon and noise sensitivity issues in High-Gain observer applications.
• To reduce the infimum of observer gain for improved state estimation.

Main Methods:

• Utilizing a Linear Parameter-Varying (LPV) / Linear Matrix Inequality (LMI) approach.
• Transforming the observer structure into a Luenberger-like configuration.
• Decomposing nonlinear system dynamics and solving adjustable LMIs.

Main Results:

• Significantly reduced observer gain infimum, mitigating peaking and noise effects.
• Demonstrated convergence using Lyapunov stability analysis.
• Effective validation on a single-link mechanical arm and vehicle trajectory estimation.

Conclusions:

• The proposed LPV/LMI method effectively reduces High-Gain observer peaking and noise sensitivity.
• The technique offers a robust solution for accurate state estimation in dynamic systems.
• Validated applicability across mechanical and automotive engineering domains.