Title :
Efficient DFT-based model reduction for continuous systems
Author_Institution :
Dept. of Electr. Eng., Louisville Univ., KY, USA
fDate :
10/1/1991 12:00:00 AM
Abstract :
A method for online identification of reduced-order models (ROM) of stable continuous systems is presented. The method is unique because it utilizes the moving discrete Fourier transform (MDFT) to continuously monitor the frequency-domain profile of the plant input and output signals. These signals need not be sinusoidal, although they must be accurately represented by their DFTs. Also, the input must contain at least n frequency components (for an nth-order ROM). A computer simulation demonstrates the method
Keywords :
control system analysis computing; fast Fourier transforms; frequency-domain analysis; identification; signal processing; computer simulation; frequency-domain profile; moving discrete Fourier transform; online identification; reduced-order models; signal processing; stable continuous systems; Computer simulation; Computerized monitoring; Condition monitoring; Continuous time systems; Discrete Fourier transforms; Discrete transforms; Fourier transforms; Frequency estimation; Read only memory; Reduced order systems;
Journal_Title :
Automatic Control, IEEE Transactions on
