Real-time IIR notch filtering using Darwinian adaption

An adaptive approach is discussed which fundamentally involves minimizing the mean-square error between a reference signal and a second-order notch filter´s output. It is pointed out, however, that practical operating implications have led to some interesting supplementary minimization constraints. A cumulative-selection procedure analogous to Darwinian evolution, in contrast to conventional deterministic and highly structured algorithms, guides the adaptation. Excellent adaptivity has been achieved, both in static interference situations and when jumps in interference frequency are imposed