RLS adaptive predictor-enhanced differentiation

The well-known noise problem in digital differentiation is addressed by means of using adaptive digital filtering for signal preprocessing. Rapidly responding differentiators with low-noise output can be constructed by using the adaptive filter in a predictor configuration. As the prefilter is adaptive, the approximation is much more flexible than polynomial fitting, for instance, and different types of signals can be processed. The RLS algorithm is used for prediction because of its good convergence properties