Research On Joint Structure Fault Diagnosis Method Based On Micro Motion Information

As the core component of the rotating machine,the joint structure plays an important role in the transmission power and connecting parts.The joint structure failure will cause the robot to work inaccurately,affecting the motion accuracy,and the joint structure failure in the crane will also affect the normal operation of the crane.Therefore,it is very practical and important to diagnose the joint structure.In this paper,the micro-sensor can not only measure the wide-scale spatial motion of the machine,but also measure the local vibration response.Therefore,it is possible to combine mechanical macroscopic motion with minute motion conditions to achieve fault diagnosis of the joint structure.Since the bearing is a rotating component that often appears in the rotating joint,mechanical failure is easy to occur,and the fault diagnosis method for the bearing has been greatly developed at present,and the research on the fault diagnosis method of the joint structure has certain reference.Therefore,this paper focuses on fault diagnosis of rolling bearings.This paper has mainly completed the following aspects:(1)This paper designed and built a joint structure failure test bench.It is mainly composed of a driving component,a transmission part,and a joint structure.The test bench realizes the relative rotation of the fixed arm and the rotating arm through the bearing,and the winding of the wire rope by the motor can realize the rotation of the rotating arm in the vertical plane.(2)This paper conducts data modeling on the experimental bench,and uses dynamic simulation software to simulate and analyze the motion of the experimental bench.By observing the simulated output data under different faults,it can be found that the acceleration data output by the experimental bench in the presence of different faults is significantly different from the angular velocity data.(3)This paper designs a nine-axis signal fusion algorithm and a trajectory fitting algorithm based on Mahoney filtering.In the signal fusion algorithm,the angular velocity signal is mainly used,and the acceleration signal and the geomagnetic signal are used to correct it.In the trajectory fitting algorithm,it is mainly the process of integrating the acceleration signal,de-drifting,re-integrating,and then drifting.The simulation data is used to verify the trajectory fitting algorithm.The fitting of the signal after high-pass filtering can simply extract the micro-motion trajectory.(4)The data processing method of the original signal is improved.Verification of the improved data processing method was found to be consistent with the ability to extract micro motion information from faulty bearings.Experiments can prove that the micro-motion trajectory can reflect the motion characteristics of different faults,different speeds and different loads.(5)Using three manifold learning dimension reduction algorithms to reduce the dimensionality of the micro motion trajectories in the three-dimensional space,and compare the advantages and disadvantages of the three algorithms on the micro-motion trajectory dimension reduction.Data processing is performed on the trajectory signal after dimension reduction,and the sensitive features are extracted by using the time domain eigenvalue method.