Attempt to Classify the Microphone Array Deconvolution Methods in Aeroacoustics
* Presenting author
In the field of experimental aeroacoustics in wind tunnel, the number of existing methods is important and is increasing continuously so that it becomes difficult for Industrials to have a clear view of the respective advantages of each of them to make appropriate choice. Benchmark works are also plentiful and often focus on the performance of these methods in terms of spatial localization and computing time, more rarely in terms of quantitative estimation. We are interested here by these deconvolution methods that we aim to classify, and even to show in what extent all of these methods are equivalent. The approach adopted is to define a common protocol for controlling and monitoring the different algorithms: define same initial conditions, same convergence parameters and same exit conditions. All of the considered deconvolution algorithms are implemented with respect to the same protocol. Convergence and calculation time are monitored at each frequency. Finally acoustic power of source area are extracted from the obtained maps and compared together. Indeed, it is observed that within the range of few dBs the results are quite similar and we propose to classify the algorithms with their equivalent Lp-norm minimization function.