Denoising of Carbon Steel Corrosion Monitoring Signal Based on NLM-VMD with MFDFA Technique

In order to recover effective information of steel atmospheric corrosion from corrosion monitoring signal, an adaptive denoising algorithm based on NLM-VMD with MFDFA technique has been processed in this paper. To eliminate the influence of outliers in the signal, NLM was used for signal preprocessing. The preprocessed signal was decomposed by VMD into several IMFs and then every IMF was analyzed by MFDFA to quantify the noise containing condition. Reconstructing the IMFs which contained none noise could obtain removed noise signal. To evaluate the performance of proposed algorithm, an experiment about artificial analog signal was carried out. The experiment result was compared with the other denoising algorithms based on mode decomposition and the proposed algorithm performed better than others. Using the proposed algorithm in corrosion monitoring signal, the quasi-periodic pulses were excellent preserved. Furthermore, this method provides the basis for research the characteristics of corrosion monitoring signal.