Characteristic Energy Ratio Ramanujan-Gram: A Novel Optimal Multibands Demodulation Method

Due to the influence of interference noise and the coupling of multistate features, it is difficult to accurately identify and extract the composite state features. In response to this issue, a method of multibands demodulation, called the characteristic energy ratio Ramanujan-gram (CERRgram), is proposed, which can extract the feature information of composite states. On the one hand, CERRgram method uses power spectral density (PSD) function instead of Fourier spectrum, which simplifies the frequency distribution of the original signal and reduces the influence of interference frequencies, thus improving the accuracy of frequency band division and providing a solid foundation for the subsequent separation and extraction of composite state features. On the other hand, the characteristic energy ratio (CER) is constructed to evaluate the features of different states. CER can evaluate the feature information of signals at various fault frequencies to identify the optimal frequency bands with rich composite state feature information. The analysis of simulation signals and experimental signals indicates that the CERRgram method demonstrates strong capabilities in dividing bands and separating and extracting composite state features.