Wiener Deconvolution for Reconstruction of Pneumatically Attenuated Pressure Signals

Wiener deconvolution is applied to the problem of reconstructing transient signals from pneumatically attenuated pressure measurements. The presented method offers a viable alternative to traditional techniques where pressures are sensed by in situ surface transducers. In situ installations present significant issues, including potential weakening of structural walls, overheating of pressure sensing elements, strain biasing of sensing elements, and resonance within the measurement port. A less complex installation involves a small surface pressure tap with the pressure signal transmitted to the transducer via a significant length of pneumatic tubing. This remote installation allows the transducer to be mounted in a controlled environment, and it virtually eliminates any sensing errors due to surface strain. The presented filter amplifies attenuated pressure signals while selectively rejecting sensor noise. The frequency-based algorithm accounts for both white and colored noise contaminations. The deconvolution method is validated using laboratory-derived step-response data. The examples presented demonstrate the algorithm accuracy for both constant and frequency-weighted signal-to-noise ratios based on the noise threshold of the measured pressure signal. Experiment results demonstrate insensitivity of the Wiener algorithm to weighting schemes used to estimate the signal-to-noise ratio of the measurements. Sensitivity analysis indicates that deconvolution fit errors are relatively insensitive to parameter perturbations near nominal measured values.