Performance comparison of the FXLMS, nonlinear FXLMS and leaky FXLMS algorithms in nonlinear active control applications

This paper investigates the performance of the filtered-X LMS (FXLMS) and leaky filtered-X LMS (LFXLMS) algorithms in a nonlinear system. In addition, we derive the nonlinear filtered-X LMS algorithm (NLFXLMS), which is obtained by considering the nonlinear system function for the determination of the gradient of the cost function. We show that the fact of not using the exact gradient implies in a performance reduction of the adaptive algorithm. The adaptive algorithms are examined in the context of active noise control applications, which possess nonlinear components. Through numerical simulations we discuss the capabilities of the examined adaptive algorithms.

[1]  José Carlos M. Bermudez,et al.  Nonlinear secondary-path effects on the transient behavior of the multiple-error FXLMS algorithm , 2000, 2000 IEEE International Symposium on Circuits and Systems. Emerging Technologies for the 21st Century. Proceedings (IEEE Cat No.00CH36353).

[2]  C. Richard Johnson,et al.  Parameter drift in LMS adaptive filters , 1986, IEEE Trans. Acoust. Speech Signal Process..

[3]  N. Bershad.  On weight update saturation nonlinearities in LMS adaptation , 1990, IEEE Trans. Acoust. Speech Signal Process..

[4]  Scott D. Snyder,et al.  Active control of vibration using a neural network , 1995, IEEE Trans. Neural Networks.

[5]  M. O. Tokhi,et al.  Active noise control systems , 1987 .

[6]  Rui Seara,et al.  Mean weight behavior of the Filtered-X LMS algorithm , 1998, Proceedings of the 1998 IEEE International Conference on Acoustics, Speech and Signal Processing, ICASSP '98 (Cat. No.98CH36181).

[7]  Rudolf Rabenstein,et al.  A measurement method for the linear and nonlinear properties of electro-acoustic transmission systems , 1998, Signal Process..