Persistently-exciting signal generation for Optimal Parameter Estimation of constrained nonlinear dynamical systems

Area of Science:

• Control Systems Engineering
• Nonlinear System Identification
• Optimization Techniques

Background:

• Accurate parameter estimation is crucial for understanding and controlling nonlinear systems.
• Traditional methods struggle with unknown true system parameters, complicating excitation signal selection.
• The need for robust methodologies that account for parameter uncertainty is evident.

Purpose of the Study:

• To develop a novel methodology for sub-optimal excitation signal generation.
• To achieve optimal parameter estimation for constrained nonlinear systems.
• To propose a benchmark-based approach for evaluating excitation signals when true parameters are unknown.

Main Methods:

• A dual-layer optimization framework is employed.
• Inner level: Nonlinear optimization minimizes output error between optimized and benchmark models for a given signal.
• Outer level: Metaheuristic optimization constructs the optimal excitation signal based on inner level fitness and experimental costs.

Main Results:

• The proposed method effectively generates excitation signals tailored for parameter estimation.
• It provides a practical approach to handle the challenge of unknown true system parameters.
• The benchmark model successfully guides the signal generation process.

Conclusions:

• The novel dual-layer optimization methodology offers a robust solution for excitation signal generation and parameter estimation in nonlinear systems.
• This approach enhances the accuracy and efficiency of system identification.
• It provides a valuable tool for researchers and engineers working with complex systems.