Research On Early Fault Diagnosis Method Of Rolling Bearing Based On VMD And ASSD

In industrial production,rolling bearing as a common part,which is very easy to fail under the action of alternating loads.Therefore,it is of great engineering significance to study the fault diagnosis method of rolling bearing.Since the fault features of rolling bearing is weak and accompanied by large noise in the early stage of failure,which makes the fault information is submerged in the noise and is difficult to extract.Therefore,it is technical difficulty to use an effective method to accurately extract the early weak fault characteristics of rolling bearing.This paper is based on the variational mode decomposition(VMD)algorithm,the continuously improves and perfects the VMD algorithm,studies the above difficulties deeply,and proposes a novel and reliable method for early weak fault feature extraction of rolling bearings.The main research contents are as follows:When the signal is decomposes by the VMD algorithm,the number of decomposition layers k and the penalty factor ? in this algorithm need to be set in advance.The selection of the number of decomposition layers k and the penalty factor ? in the VMD algorithm usually adopts trial-and-error method and single-objective optimization method.But the above method cannot obtain the optimal parameter combination,so that the VMD algorithm does not have adaptive.Therefore,this paper proposes a multi-objective particle swarm optimization(MOPSO)algorithm to optimize the parameters of VMD.The steps are: Firstly,the symbol dynamic entropy(SDE)and power spectral entropy(PSE)are selected as fitness functions of the MOPSO algorithm;Then use the MOPSO algorithm to optimize the iterative process,and obtain the Pareto optimal frontal solution set;Finally,the optimal combination of VMD parameters [k,?] is obtained by normalization.In the study of VMD algorithm,it is found that noise has a certain impact on this algorithm.In the strong noise environment,it is not ideal to use VMD method to extract fault features of rolling bearing.At this time,it is necessary to perform noise reduction preprocessing on the signal containing strong noise,and then use the VMD method to extract the fault feature.Based on the above views,the sparse-spike deconvolution(SSD)algorithm is introduced in this paper.In SSD algorithm,L1 norm regularization method is used to sparse the constrained signal pulse sequence,which makes the result sparse.However,considering that the L1 norm regularization parameter is not adaptive,which makes the signal processing result not ideal.This paper proposes a method to use the quantum-behavior particle swarm optimization(QPSO)algorithm to determine the optimal L1 norm regularization parameter adaptively.This makes the sparse-spike deconvolution algorithm is adaptive,that is,the adaptive sparse-spike deconvolution(ASSD)algorithm,and applies this algorithm to the noise reduction of early weak fault signals of rolling bearings.Based on the continuous improvement and perfection of the SSD algorithm and the VMD algorithm,this paper proposes a novel early weak fault diagnosis method of rolling bearing based on ASSD and VMD.At the same time,this paper uses effective simulation signals and experimental signals for analysis to verify the effectiveness of the proposed method.