System Excitation Optimization for Accurate Model-Based Control

The patent transforms the system identification approach from relying on fixed empirical measurement catalogs to using systematically generated excitation signals with optimized parameters. By changing the parameters of excitation signals (frequency, amplitude, duration) based on system characteristics rather than relying on predefined empirical data, the method achieves both accuracy and adaptability to new technologies.

The patent replaces the expert-knowledge-based empirical system with a systematic, algorithm-based approach for generating and optimizing excitation signals. This substitution eliminates dependency on empirical measurement catalogs and expert experience, enabling the method to work effectively with new technologies where such empirical data is unavailable.

The patent applies preliminary action by systematically designing and optimizing excitation signals before conducting system identification. By pre-determining the optimal characteristics of excitation signals based on system models and identification objectives, the method ensures high model accuracy without requiring complex trial-and-error adjustments during implementation.

The patent introduces dynamics by making excitation signal characteristics adjustable and optimizable based on specific system identification needs. Rather than using fixed or simple excitation signals, the system dynamically adapts signal parameters (frequency content, amplitude modulation, time duration) to match the requirements for identifying specific system characteristics, thereby achieving high accuracy without excessive complexity.