Hybrid algorithm for adaptive IIR notch filter

Abstract

In this paper presents the hybrid algorithm for adaptive IIR notch filter using the combination of genetic algorithm in parallel with variable step-size algorithm. This application is used for frequency estimation in background noise such as Gaussian noise or impulse noise. The objectives are enhancing the performance of adaptive IIR notch filter, for instance, fast convergence speed, low mean square error (MSE) and especially for noise robustness in high amplitude impulse noise situation. From the experiment results show that the proposed algorithm gives good performance, such as fast convergence speed, noise robustness, low MSE and can be used to eliminating 50 Hz power line interference, which induce to disturb the signal of electrocardiogram (ECG) of patient while the physician is recording it.