Applications of wavelet transforms to system identification

From MaRDI portal





The authors first give a review of system identification based on distribution theory. Using the Schwartz kernel theorem it is possible to associate a unique distribution (called kernel distribution) to any continuous linear system. Then formulae using wavelet transform to access time-frequency information of the kernel distribution are deduced. By using the denoising property of wavelets the authors prove that their identification method works better than conventional techniques when white noise is added to input and output. Numerical experiments are given. For the entire collection see [Zbl 1050.47001].











This page was built for publication: Applications of wavelet transforms to system identification

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4659382)