Research On Fault Diagnosis Method Of Wind Turbine Bearing Based On Probability Box Theory And HGWO-SVM

At present,the development and utilization of clean energy is one of the key research topics in the world.Wind energy has attracted much attention because of its large reserves,wide distribution and high utilization value.Wind energy is gradually converted into electricity needed by industry and daily life through the operation of wind turbines,and the installed capacity of wind turbines continues to grow steadily every year,which has broad prospects for development.However,the wind turbine is located in a remote location,the working environment is often poor,prone to accidents,affecting the safe and stable operation of wind turbine.Bearing as a key rotating component of wind turbine,its failure rate is often high,once a fault occurs,it can cause shutdown,reduce power generation and economic benefits.It can be seen that the health of the bearing has a great impact on the normal operation of the whole wind turbine.Therefore,in order to prevent the major outage accident and reduce the unit operation cost,this paper takes the rolling bearing with high fault incidence in wind turbine as the diagnosis object,analyzes and studies the bearing fault feature extraction method and fault diagnosis method.Combined with the relevant theoretical basis of probability box and support vector machine,an intelligent mechanical fault diagnosis method is proposed.First of all,according to the components and working principle of wind turbine,the types,characteristics,common faults and fault forms of wind turbine bearings are further analyzed,and the fault diagnosis scheme design is completed on the basis of the existing literature.Aiming at the uncertainty problems such as information loss and misoperation of rolling bearing fault vibration signal in feature extraction and the unsatisfactory accuracy of fault diagnosis,the probability box theory is introduced.The probability limit analysis is carried out by using the public bearing data of case Western Reserve University,and the cumulative uncertainty measurement method is used to extract the direct modeling method to construct the probability box of fault signal.The eigenvector of the probability box is used for the data input of support vector machine fault classification,which lays a foundation for the fault identification and classification of wind turbine bearings.Secondly,in order to improve and optimize the performance of support vector machine parameters,in view of the shortcomings of the original grey wolf optimization algorithm,nonlinear convergence factor,Levy flight strategy and greedy retention strategy are introduced to improve it.an algorithm for parameter optimization of support vector machine is designed,which provides an effective algorithm tool for subsequent parameter optimization of support vector machine.Finally,the basic theory of statistics is studied,the mathematical model of support vector machine is deduced in detail,and the improved grey wolf optimization algorithm is used to optimize the two parameters C and σ of support vector machine,so as to establish the intelligent fault diagnosis model of bearing.combined with the extracted wind turbine bearing probability box features,the faults in different positions of bearings are diagnosed and accurately identified and classified.To a certain extent,it avoids the errors caused by expert experience or accidental factors,which makes it more applicable in various engineering fields and more efficient in bearing fault diagnosis.The method proposed in this paper makes full use of the outstanding advantages of probability box in dealing with uncertainty and the excellent classification performance of support vector machine in accurate recognition with only a small number of samples.this has no doubt in the accurate identification of vibration signals of different fault types.The experimental results show that the intelligent fault diagnosis model established in this paper is feasible and effective in bearing fault diagnosis of wind turbine.