LAUSR

Sample entropy (SampEn) technologies have been widely applied in diagnosing the faults of industrial systems. However, there are two disadvantages of these technologies. First, all of these technologies measure the distance of two vectors solely based on the maximum distance between the corresponding elements in the two vectors, which is not able to fully reflect the distance of the two vectors. Second, these methodologies measure the similarity of two vectors with either zero or one, which may cause sudden changes in entropy values. Therefore, we proposed a Euclidean distance based multiscale fuzzy entropy (EDM-Fuzzy), which measures the similarity of two vectors with continuous values from zero to one based on the Euclidean distance of the two vectors. The results from the synthetic and real signals demonstrated that EDM-Fuzzy has higher accuracy in measuring the complexity of signals. As a result, EDM-Fuzzy obtains a higher accuracy in detecting the bearing faults than the state-of-the-art entropy technologies.