Reserach On Early Weak Fault Diagnosis Of Rolling Bearing Based On Duffing System

In the field of machining,rolling bearings are extremely widely used.Once a rolling bearing fails,it will affect the operational reliability of the entire equipment.In the early stage of the fault,its signal characteristics are very weak and difficult to identify.Weak faults will develop into obvious fault characteristics after a period of time,which will seriously affect production safety and cause great harm.This article starts research on how to identify early weak faults.The thesis first studied the method of detecting the periodic signal using the traditional Holmes-type Duffing system,and the Jacobian method to calculate the Lyapunov exponent of the Duffing system.The chaotic threshold of the Duffing system was obtained.Aiming at the limitations and deficiencies of traditional systems that cannot detect high-frequency signals and require frequent calculation of chaos thresholds,the system is improved by using a generalized time-scale transformation method,so that the frequency range of the system detection signals is expanded.The chaos threshold is recalculated each time the detection is performed,making the detection process more efficient.Through derivation analysis,when the phase difference between the periodic built-in driving force of the Duffing system and the periodic test signal of the input system is between90° and 270°,the system will produce misjudgment.Aiming at this problem,the method of adding initial phase angle to change the built-in driving force of the system and the phase shift method to change the phase of the signal to be measured are proposed,which can effectively solve the problem of misjudgment.In order to better ensure the integrity of the signal to be measured,the method of adding the initial phase angle is selected to solve the problem of misjudgment,and the improved Duffing system model is finally obtained.When the Duffing system is in a large-period state,the variance of its system displacement output value is greater than that of the system in chaotic state,and the output value variance canobtain the maximum value when the frequency of the system is consistent with the signal to be measured.By using this rule,the identification of weak periodic signals is realized.When using the improved Duffing system to identify the signal to be measured,two or more calculations are needed to generate multiple displacement output value variances.The output value variances are reconstructed for this phenomenon,so that the Duffing system can find the output value variances.Maximum value to identify weak periodic signals.Due to the complex noise prevailing in the measured signals,a simulated signal containing colored noise was prepared by filtering the white noise,and the simulated signal was used to stimulate the system to successfully identify the weak fault signal to be tested,which proved The detection method in this paper is reliable.Then the paper uses the measured weak early fault signal of the rolling bearing excitation system to successfully identify the fault signal.A method for diagnosing the early weak faults of rolling bearings based on the Duffing system is proposed.After a large amount of data verification,the method is proved to be effective for the identification of early weak fault signals.This method provides a new idea for the identification of early weak faults.