Automatic oscillation detection based on improved local mean decomposition

A novel method based on improved local mean decomposition (LMD) for oscillation detection is proposed. The local mean decomposition, which is of particular relevance with regard to analyze amplitude and frequency modulated (AM-FM) signals, is especially suitable for characterizing oscillations in measurements from control loops in industrial processes. In contrast to empirical mode decomposition (EMD), the improved LMD performs better in the following aspects: (i) ability to extract both single/multiple oscillations in short time series, (ii) robustness to noise, and (iii) capability to handle time-varying oscillations. Simulation examples and industrial application are provided to demonstrate the effectiveness of the proposed LMD oscillation detector.