Fault Diagnosis Of Spindle Bearing Of High Concentration Disc Refiner

High concentration disc refiner is an important grinding equipment in papermaking industry.The stable ability of grinding paper pulp of the machine directly affects the economic benefit of enterprises.The spindle of high concentration disc refiner belongs to precision spindle category,and the minimum distance of the fine grinding area between the grinding discs is less than 0.1mm.If the spindle system cannot work stably,the moving disc will cause the collision of the grinding disc under the condition of deflection,and the high concentration disc refiner will be damage greatly by the occurrence of "collision disc".As a crutial part of the spindle system of high concentration disc refiner in bearing loads,the spindle bearing is liable to fail under the impact of alternating load during operation,thus affecting the normal operation of high concentration disc refiner.Therefore,it is necessary to study the fault diagnosis of spindle bearing of high concentration disc refiner.Firstly,the working characteristics of spindle bearing of high concentration disc refiner are analyzed,and the components of vibration signal of high concentration disc refiner spindle bearing fault are analyzed.Aiming at the problem that the vibration noise of high concentration disc refiner can cover the fault information under the actual working conditions,the multipoint optimal minimum entropy deconvolution adjusted mothed is introduced to denoise the original vibration signal and strengthen the periodic impulse shock characteristics.The bearing fault test bench is built to collect fault data of rolling bearings.In order to simulate the actual working conditions of the high concentration disc refiner,add noise to the fault signals collected,the effects of the noise decrease and the features increase of simulation signal of high concentration disc refiner spindle bearing fault are verified by contrast experiments.It provides an important guarantee for realizing fault features extraction of spindle bearing of high concentration disc refiner.Secondly,a variational modal decomposition method based on chaos modified jellyfish search optimization algorithm is proposed to realize adaptive extraction of fault features of spindle bearing of high concentration disc refiner.The ability of variational mode decomposition,empirical mode decomposition and local mean decomposition to extract signal characteristics is verified through comparative experiments.Aiming at the defect that the number of decomposition and penalty coefficient need to be set artificially in variational modal decomposition,jellyfish search optimization algorithm is introduced to optimize the variational modal decomposition,and chaos mapping is improved to increase its capability,so as to realize the adaptive parameter determination and decomposition of different types of fault signals.The feature extraction capability of the proposed method is verified by comparative experiments.Finally,a dense convolution network based on spatial attention recalibration is proposed to classify the faults of high concentration disc refiner spindle bearings.In order to solve the traditional artificial identification of high concentration disc refiner spindle bearing fault requires abundant expert experience and the key information in data cannot be effectively extracted,densely connected convolutional network is introduced to accurately identify rolling bearing faults.The spatial attention mechanism is introduced to form a densely connected convolutional network for spatial attention recalibration,and two compression modules are designed to compress the model.In the actual working condition,the vibration signal of the spindle bearing of the high concentration disc refiner is collected,and the fault diagnosis of the spindle bearing of the high concentration disc refiner is realized through the fault diagnosis method proposed,which has practical significance to ensure the stable operation of the high concentration disc refiner.