Font Size: a A A

Research On Gas Turbine Fault Diagnosis Method Based On Analytic Hierarchy Process

Posted on:2015-01-14Degree:MasterType:Thesis
Country:ChinaCandidate:Y D LiFull Text:PDF
GTID:2132330434457974Subject:Precision instruments and machinery
Abstract/Summary:PDF Full Text Request
As the power equipment of gas turbine set, gas turbine easily has hidden faults in the running process because of severe work environment, peak shaving needs, and service wear and so on. Improve and perfect the fault diagnosis system architecture of gas turbine and timely do the fault diagnosis to gas turbine are good for ruling out the hidden dangers, reducing accident rate and improving the safety, which provide the theory evidence for the safe operation of the set and the improvement of economic returns. Therefore, the study of fault diagnosis methods of gas turbine has great practical significance in reliable operation, security insurance, service life extending, and scientific management and maintenance etc.In this paper, QD100gas turbine, used by CNPC-Turkmenistan to generate electricity, is made the study object of fault diagnosis methods of gas turbine. The main study contents include:analyze the mechanisms, modes of typical gas turbine faults, and analyze the mapping criteria and relationships among the fault features; Based on the system structure characteristics of gas turbine, analyze the gas turbine fault tree and use AHP to analyze the typical gas turbine fault---compressor start fault, and give repair advice; With the hierarchical analysis of main fault mode and fault tree of gas turbine, construct a comprehensive evaluation system of the gas turbine state, and use the system to evaluate the state of gas turbine in the standard conditions; Adopt the method of data mining based on rough set theory, do attribute reduction to fault samples to simplify the dimensions. And combining rough set theory with the neural network method, offer a proposal of applying it to gas turbine fault diagnosis; Apply all the study contents to gas turbine fault diagnosis system, add the section of comprehensive evaluation of gas turbine state and improve and perfect the functions of gas turbine fault diagnosis system in MATLAB and C#.The results of gas turbine fault events analysis based on fault tree indicate its clear effect to fault events hierarchy relationship. AHP is used in comprehensive evaluation system of gas turbine state to evaluate the index system, whose evaluation effect gets a validation in evaluating the gas turbine state under standard operating conditions. Therefore, the evaluation results of comprehensive evaluating index system of gas turbine state can be the grounds of further diagnosis and maintenance plan. The test of improved fault diagnosis software system of gas turbine, which is based on AHP assessment system, shows that the evaluation results are good for the operators to master the working performance according to the evaluation of operation state, which could be the grounds source of diagnosis and maintenance plan.
Keywords/Search Tags:Gas turbine, Fault diagnosis, AHP, Rough set, Neural networks
PDF Full Text Request
Related items