Denoising sensors signals within mechatronic systems using a Wavelet technique

In many real world applications, signals are nonstationary. One solution for processing non-stationary signals is the Wavelet Transform. Currently, there is tremendous focus on the application of Wavelet Transforms for real-time signal processing. This leads to the demand for efficient architectures for the implementation of Wavelet Transforms. In this paper, a Wavelet shrinkage technique is applied to denoise three real signals from the sensors of a test bench for aeronautical equipment. First, we discribe different architectures of Wavelet Transform for real-time signal processing. Then, these architectures are implimented in Simulink environment. Two thresholds are used, the universal threshold and a new threshold is proposed for improving the result of denoising. At the end of this article we discuss and compare different denoising results obtained with the universal and our proposed thresholds using FFT spectrum and MSE (Mean Square Error) as criteria for evaluating performances results.