Compressive sensing based spinning mode detection with in-duct microphone array
Wenjun Yu, Xun Cheng Huang · 2016
A compressive sensing based mode detection method for duct spinning mode is presented in this paper. This method could break through the spatial Shannon-Nyquist sampling theorem and use less sensors than conventional mode detection method when the incident duct mode is sparse at blade passing frequencies, and save the costs of data acquisition, signal processing and data storage in experiments. Numerical simulations validate the effectiveness of this method and evaluated the probability of successful detection with different number of selected sensors and mode sparsity. This method could successfully detect the main modes when the signal-noise ratio SNR > 20 dB. A circular duct with spinning mode synthesizer and microphone array was adopted to validate the performance of this compressive sensing method in experiment, and the test results shows the compressive sensing method could use less than half of the sensors to get a satisfied mode detection spectrum. This new method also have the ability of detecting higher order mode with the same minimum spacing between sensors, and this advantage is capable to improve the mode detection ability of higher order modes for existing facilities.