A robust data-driven system identification approach with applications to reliability analysis of DC motor systems

Area of Science:

• Electrical Engineering
• Control Systems
• Reliability Engineering

Background:

• System disturbances are critical in reliability analysis, especially in DC motor systems.
• Unknown periodic disturbances pose significant challenges for accurate system identification and reliability assessment.

Purpose of the Study:

• To propose a robust, data-driven system identification approach for DC motor reliability analysis.
• To effectively identify and eliminate unknown periodic disturbances impacting system performance.

Main Methods:

• Utilizing subspace methods and modified Bernstein polynomials to construct a disturbance space.
• Developing a data-driven closed-loop input-output (I/O) extended model.
• Employing orthogonal projection techniques to eliminate disturbances and canonical correlation analysis (CCA) for parameter estimation.

Main Results:

• Successfully designed a scheme for unknown periodic disturbance space construction.
• Derived a closed-loop I/O extended model incorporating controller information.
• Demonstrated effective elimination of disturbances and accurate system parameter estimation.

Conclusions:

• The proposed data-driven approach provides a robust solution for reliability analysis of DC motor systems with unknown periodic disturbances.
• The method's feasibility and effectiveness are validated through a DC motor system case study.