Active control is a methodology based on the waves destructive interference that has proven to be effective for attenuating noise in the low frequency audible spectral range. However, the case of impulsive type noise sources, as harmful as frequent in industrial environments, represents a challenge to the convergence of the control algorithm that is still a matter of study. Outliers in the measured signals cause overcorrections in adaptive adjustment of filter weights which can produce instability. This paper presents the results of applying a new robust methodology to attenuate impulsive noise in a single-channel system. The proposed algorithm based on the maximum correntropy criterion with recursively adjusted kernel size, does not require prior statistical information on noise. The convergence properties and the effectiveness of the control indices are verified by simulation in different conditions of noise environments. Impulsive noise is represented by the nongaussian model proposed in the bibliography.Workshop: WPSSTR - Procesamientos de Señales Sistemas de Tiempo RealRed de Universidades con Carreras en Informática