Reliability Analysis Of Compressor Rotor System In Gas Turbine Based On Bayesian Network

With the development of society,the energy issue has attracted more and mor e attention.The country has also put forward the urgent need to adjust the energy structure.High efficiency,energy saving,and emission reduct ion have become the inevitable development trend of China’s power generation industry.Gas turbines have high power generation efficiency and low pollution.At the same time,with the improvement of combustible ice and shale gas mining technology,gas turb ines and gas-steam combined cycles will be an important part of future power production.However,the operating conditions of gas turbines are complex,and there are many factors that affect its operational stability,reliability,and safety.In order to e nsure the efficient and safe operation of the gas turbine,this paper takes the gas turbine compressor rotor system as the research object and carries out the reliability analysis.The main work of this article are:1.Considering the shortcomings of the traditional Bayesian network model in reliability evaluation,the theory of common cause failure,the theory of multi-state system and the theory of grey system were introduced into the BN.The parameter-model and structure-model of traditional BN were extended,and a new model named “Grey Bayesian network reliability analysis model for multi-state systems based on common cause failure” has been established,which improves the Bayesian network’s ability of solving uncertainty problems.2.Taking the gas turbine compressor rotor system as the research object,the functional structure,failure mode and failure cause of the compressor rotor system were analyzed in detail,and the fault tree model of the compressor rotor system was established.Through the method of model conversion,the Bayesian network reliability model of the compressor rotor system was established,and the reliability analysis of the rotor system qualitative and quantitative was completed.Rotor blade fouling and corrosion are the weakest link in the rotor system.3.Based on the traditional two-state assumption,the compressor rotor system was multi-state expanded and the possible common cause failures in the system were considered.The interval gray number was used to describe the uncertain f ault information and logical relationship between components,and multi-state grey reliability model based on Bayesian network of the compressor rotor system was established.Through the model calculation,the grey reliability of the compressor rotor system was obtained,and by comparing with the calculation results of the traditional Bayesian network model,the correctness and availability of the established model were proved.Reliability evaluation problem of mechanical system in the presence of multi-state and common-cause failure could be solved.4.Combining the reliability analysis results of the compressor rotor system with the actual engineering status of the gas turbine power plant,in view of the current deficiencies of the gas turbine power plant in handling fouling problems,it was proposed to establish an on-line gas turbine performance monitoring system integrating "on-line particle capture detection system" and "on-line gas turbine efficiency monitoring system".This system can solve the problem of monitoring and control of the fouling of the gas turbine.