Role of entropy in fault diagnosis of mechanical equipment: a review

Abstract

Abstract Mechanical equipment is widely used in daily life and production manufacturing, and it is an indispensable part of modern society. Fault diagnosis of mechanical equipment can effectively diminish catastrophic failures and significant economic losses. How to evaluate their status is the key problem of safe operation. The monitoring data of mechanical equipment obtained by sensors can be used to analyze potential problems and fault information. The vibration signal is the most easily obtained and commonly used monitoring data of mechanical equipment. But the vibration signal is unstable and usually shows nonlinear characteristics in actual measurement, which makes it difficult to extract fault features. Entropy can quantify the complexity of time series and detect the dynamic change of nonlinear behavior. Relying on the performance of entropy, it can be used as an effective tool for dynamic characteristics and applied to the fault diagnosis of mechanical equipment. The purpose of this paper is to summarize the related research of mechanical equipment fault diagnosis based on entropy methods in recent five years. Different entropy methods are classified, and specific application methods are introduced. On this basis, these entropy methods are deeply discussed. The gaps filled by different entropy methods in mechanical equipment fault diagnosis are discussed. The problems faced by entropy methods in the application of mechanical equipment fault diagnosis are also discussed. Theoretical and engineering guidance of mechanical fault diagnosis research is provided, which is of great significance.