Robust algorithm for estimation of time-varying transfer functions

We introduce a new method to estimate time-varying transfer functions and time-varying impulse response functions. The method is based on time-varying autoregressive moving average models in which the time-varying parameters are accurately obtained using the time-varying optimal parameter search (TVOPS) algorithm which we have developed. The TVOPS is more accurate than the recursive least squares, and remains robust even in the case of significant noise contamination. Application of the new method to renal blood pressure and flow revealed that hypertensive rats undergo more complex and time-varying autoregulation in maintaining stable blood flow than do normotensive rats. The newly-developed approach may promote the broader use of time-varying system identification in studies of physiological systems.