Font Size: a A A

Vibration Signal Fault Diagnosis Based On Quantum-behaved Particle Swarm Optimization Algorithm

Posted on:2018-08-16Degree:MasterType:Thesis
Country:ChinaCandidate:A H ZhuFull Text:PDF
GTID:2322330515969179Subject:Mechanical engineering
Abstract/Summary:PDF Full Text Request
In the contemporary plant,the mechanical system is becoming more and more continuous high-speed and large-scale.In order to minimize the accidents and losses caused by equipment failure,it is necessary to monitor production equipment in real time.It is of great significance.In industry,vibration signals are ubiquitous and easy to collect.Therefore,It is of great significance to diagnose the equipments based on the vibration signal of the device.Among the mrthods of reducing signal noise and fault diagnosis,the commonly used methods that wavelet threshold filtering noise reduction and fault diagnosis based on Elman neural network were studied in this paper.And a method that using the Global search optimization capability of Quantum-behaved Particle Swarm Optimization(QPSO)algorithm to optimize the defect of wavelet threshold filtering noise reduction and Elman neural network was come up.In the process of signal noise reduction by using wavelet threshold filtering method,the size of the threshold determines the quality of the signal which has been noise reduction.Therefore,using QPSO algorithm to optimize the wavelet threshold filtering was proposed.The optimal threshold could be found based on global search optimization ability of QPSO algorithm.Then,signal noise reduction could be performed.After simulation,the results showed that the effect of the proposed noise reduction filtering method was better than wavelet threshold filtering method.For some defect of Elman neural network,such as easy to fall into the local minimum,this paper decided to use QPSO algorithm to optimize it.Some of the defects in the Elman neural network were eliminated based on the global search optimization performance of QPSO algorithmAt the end,the correctness of.the proposed method was verified by the vibration signal data of rolling bearing.Wavelet threshold filtering which optimized by QPSO algorithm was used to filter the vibration signal of the rolling bearing.And the result was compared with the general wavelet threshold filter,The result showed that the proposed method had achieved better result.Then,the Elman neural network optimized by QPSO was verified.At the same time,the result was compared with Elman neural network.The result showed that the Elman neural network which optimized by QPSO algorithm could quickly reach the default error target value and had a higher recognition rate.Therefore,the proposed method was of feasible,effective and accurate.The fault diagnosis method that optimized by QPSO algorithm had great advantage and application prospect.
Keywords/Search Tags:Vibration signal, fault diagnosis, QPSO, wavelet threshold filtering, Elman neural network
PDF Full Text Request
Related items