Precise measurement of mechanical forces is crucial\nto efficient micro-manufacturing. The quality of such\nmeasurements depends heavily on the properties of the\nnoise inevitably accompanying every measurement process.\nIn the micro-range, the signal-to-noise ratio tends to be very\nlow, and the noise dynamic varies for different frequencies.\nIn result, common denoising methods that assume white\nnoise perform poorly in this setting. In this paper, a novel,\neasily implementable denoising method based on a local\nstatistic of the measured data�s spectrum is proposed. By\ntesting it on a representative dataset, it is shown that the\nproposed method is robust and stable. Particularly, it allows\nfor an efficient retrieval of the force signal encountered in\nmicro-milling processes.
Loading....