Virginia Tech. Department of Mechanical EngineeringDuke University. Department of Mechanical Engineering and Materials ScienceClark, Robert L.Gibbs, Gary P.2015-05-272015-05-271994-10-01Clark, R. L., & Gibbs, G. P. (1994). Implications of noise on the reference signal in feedforward control of harmonic disturbances. Journal of the Acoustical Society of America, 96(4), 2585-2588. doi: 10.1121/1.4100650001-4966http://hdl.handle.net/10919/52658The purpose of this Letter is to detail the effect of random (zero-mean) noise on the reference input of an adaptive feedforward filter in the control of single-frequency disturbance inputs. Results from this study indicate that if system identification is required to execute the adaptive algorithm (as is the case with the filtered-x LMS algorithm), then the sample rate should be set at 4 samples per period to minimize the number of finite impulse response filter coefficients needed to compute an accurate estimate of the filtered reference signal for multi-channel control. However, results from this study demonstrate that the sample rate has very little effect on the minimum mean-squared error that can be achieved with the adaptive control filter, owing to the reverberant characteristics of the dynamics system.4 pagesapplication/pdfen-USIn CopyrightRandom noiseArticle - Refereedhttp://scitation.aip.org/content/asa/journal/jasa/96/4/10.1121/1.410065https://doi.org/10.1121/1.410065