A preconditioned LMS algorithm for rapid adaptation of feedforward controllers

A form of LMS algorithm for adaptive feedforward control is presented in which the physical plant response is preconditioned by the inverse of its minimum phase part. The overall response from the output of the control filter to the output of the error sensors is thus equal to the all-pass part of the plant response. Since this has a flat magnitude response, the convergence speed of the adaptive algorithm is not affected by resonances in the plant response, as it is for the normal filtered reference and filtered error LMS for example. Simulations are presented for a control system using a plant response measured from a loudspeaker to a microphone inside a car which support these observations.

[1]  Stephen J. Elliott.  FILTERED REFERENCE AND FILTERED ERROR LMS ALGORITHMS FOR ADAPTIVE FEEDFORWARD CONTROL , 1998 .

[2]  John Garas,et al.  Using phase information to decorrelate the filtered-x algorithm , 1998, Proceedings of the 1998 IEEE International Conference on Acoustics, Speech and Signal Processing, ICASSP '98 (Cat. No.98CH36181).

[3]  Eric A. Wan,et al.  Adjoint LMS: an efficient alternative to the filtered-x LMS and multiple error LMS algorithms , 1996, 1996 IEEE International Conference on Acoustics, Speech, and Signal Processing Conference Proceedings.

[4]  Philip A. Nelson,et al.  Active control of vibration, 1st edition , 1996 .

[5]  Stephen J. Elliott,et al.  Optimal controllers and adaptive controllers for multichannel feedforward control of stochastic disturbances , 2000, IEEE Trans. Signal Process..

[6]  Bernard Widrow,et al.  Adaptive Signal Processing , 1985 .