A NOVEL MULTIPLE-REFERENCE, MULTIPLE-CHANNEL, NORMALIZED FILTERED-X LMS ALGORITHM FOR ACTIVE CONTROL OF PROPELLER-INDUCED NOISE IN AIRCRAFT CABINS

摘要

The results show that the actuator-individual FX LMS algorithm works well when the correlation matrices are diagonally dominant. The noise reduction obtained is significant and the results indicate that the algorithm shows a good potential for achieving agreement with the optimum noise reduction (least squares solution). The actuator-individual FX LMS algorithm constitutes a good approximation of Newton's algorithm. The evaluation results also show that the actuator-individual FX LMS algorithm has better tracking performance than the global normalized and reference-individual FX LMS algorithms.