The Full Vector-gray Method For The Prediction Of The Spectral Structure Of Equipment

Equipment failure prediction is to through technical means,predict a certain time in the future operation of the unit,foresee the equipment failure what will happen,in order to guide the production scheduling and equipment maintenance.At present,the domestic and foreign scholars for fault prediction research are mainly focused on the failure strength,residual life,and trouble free or not,implement most in advance of a certain single fault,and for the nature and types of fault prediction is relatively small.To forecast fault types and properties is need to forecast the development trend of the frequency spectrum structure of equipment vibration signals,and according to the prediction of the proceeds of the frequency spectrum structure determine the fault nature and type of unit.When vibration signals which collected from different directions were proceeded spectrum analysis,found that the frequency spectrum structure exists difference,therefore,only rely on monophyletic signal for prediction is difficult to reflect the operation state of the unit.In order to improve the reliability of rotating machinery fault prediction,the technology based on full vector spectrum of homologous information fusion technology is used into the prediction model,which can effectively avoid the monophyletic signal fault feature extraction is not complete,complete statement of the rotor within the whole cross section of vibrational state.The main research work and result are as follows:(1)Research on the frequency spectrum prediction method of full vector-GM(1,1)model and its modeling process.Show the specific modeling process of the trend prediction method combining grey GM(1,1)model of full vector-GM(1,1)model of full vector spectrum technology.Tests show that the vector-GM(1,1)model has higher prediction accuracy,and it shows a good prediction effect in mechanical vibration intensity forecast.(2)Research on the frequency spectrum structure prediction process based on the full vector-MGM(1,m)model and its application.The MGM(1,m)model of grey prediction method combined with full vector spectrum technology is applied to mechanical vibration spectrum prediction research,and the whole process of the spectral prediction is given.The results show that the full vector-MGM(1,m)model has better prediction effect,and it shows a higher prediction precision in the medium and long term forecast.(3)Research on the frequency spectrum prediction process based on the full vector-BPMG(1,1).Using the EMD method to separate trend quantity and random quantity in characteristic value.Trend quantity will put into the GM(1,1)model to forecast and random quantity will put into BP neural network prediction,and finally get the predicted quantity through integrating the predicted results.The experimental results show that the full vector-BPGM(1,1)model has higher prediction accuracy and can provide technical support for predicting maintenance equipment.