Approximate Entropy Based on Hilbert Transform and its Application in Bearing Fault Diagnosis

A method of bearing fault diagnosis based on approximate entropy and hilbert transform is proposed. The main idea is: the original vibration signal is transformed by hilbert transform firstly, then calculating the approximate entropy value of the transformed signal. By studying the approximate entropy under different fault sizes of bearing inner, the regularity of approximate entropy change under the above conditions is obtained, and the feasibility of approximate entropy in bearing fault diagnosis is further confirmed.