Research On Technology Of Gas Path Fault Diagnosis Of Gas Turbine

Gas turbine work is a process that is a cross-coupling of the thermodynamic; it is proneto occur various failures under poor working conditions. It will bring great losses toproduction and operation in the event of failure. Gas turbine fault diagnosis can help peopleof operation and maintenance advance which is abnormal. We should identify the cause of themalfunction quickly and give a reasonable solution, also, make the gas turbine work safelyand steadily. In this work, this article researches on technology of gas path fault diagnosis ofgas turbine which provides a reference for the actual production. The main work is following.(1) Establishment of a standard thermal parameter model. According to the dataprovided by the manufacturer of the test, we use the calculation of variable heat cycle toreceive a characteristic line of gas turbine components. We set up a single-axis model for gasturbine fault-free by modular modeling method; use the conditions of the design point toverify the correctness of the model. It can be achieved on the monitoring of gas turbineperformance parameters through no fault model. Also, Trouble-free gas path laid thefoundation for gas path fault diagnosis of gas turbine.(2) Research on technology of gas path fault diagnosis of gas turbine. We make thedeviation of the fault model least and isolate the independent variation of changes in theamount of deviation. We associated the deviation of the measured parameters andperformance parameters by fault coefficient matrix. It is based on the criteria of the commongas path fault; we obtained the nine deviation fault situation of the typical way. Finally, weanalyzed the impact of the environmental temperature that was on the gas path diagnostic.(3) Research on machine learning application of gas path fault diagnosis of gas turbine.This paper describes the architectures and principles of BP neural network andsupport vector machine when using in gas path fault diagnosis of gas turbine. Weestablished the fault diagnosis system of gas turbine by using BP neural network and supportvector machine based on characteristic parameters of the design point.(4) The development of a framework for condition monitoring and fault diagnosis ofsingle shaft gas turbine. This paper has developed a framework of the gas turbine remote monitoring and fault diagnosis system based on NI-Labview. The system consists of dataacquisition, data storage, condition monitoring, fault diagnosis expert system in the ring,strong management capabilities, and easy expansion. It can provide support for gas turbineworking safely and steadily.