A Novel Prognostic Grey Model And The Application To The Health Management Of Gas Turbine

The appropriate maintenance strategy can make a sustainable,efficient and safe operation of gas turbine,which will have a great effect on the productivity and competitiveness of organizations and companies.For a long period,people have been working hard against the increasing com-plexity of the gas turbine health management.Due to the equipment grow-ing increasingly complex with applications increasingly widespread,equipment failure could not be eliminated completely,and the traditional maintenance method incompetent to eliminate fault of equipment and sys-tems for proper maintenance tasks.As a result,the condition-based maintenance(CBM)is widely used in the complex gas turbine maintenance.The uncertainty and failure of the integrated equipment can be determined by the condition maintenance,and the health management of the equipment can be implemented through the state monitoring,diagnosis and prognosis.The predictive maintenance(PdM)with the modernization of online mon-itoring technology is becoming increasingly mature,which is becoming the most advanced maintenance method.In order to make PdM possible,a high precision prediction method is the key to the problem.Grey model is commonly used for prediction of system with partly unknown structure,parameters and characteristics and gas turbine gas path scale process with degradation,which not only has to determine the growth trend but also the random fluctuation.But there are still need to be im-proved.In this paper,the single sequence grey prediction model and grey relational analysis combined guidance information for the reference se-quence is introduced for the prediction and combined with Grey Markov model of the structure,to make prediction model can predict the volatility.In order to complete the mature PdM scheme,this paper on the new pre-diction model promotion,a new reliability function is defined,combined with non-fully renewed recovery,unrecoverable fault residual characteris-tics,early failure characteristics,to meet certain availability based on maintaining the economy as the goal,a gas turbine gas path washing scheduling model,and the current timing maintenance model are compared and demonstrated the model economy.Combining prediction and schedul-ing of the two aspects,demonstrates the predictive maintenance of PdM in the gas turbine operation of the advanced nature of the practice.