Reliability Analysis And Optimization For Domestic Control System Of Heavy-Duty Gas Turbine

The domestic control system of heavy-duty gas turbine is one of the major strategic objectives of China’s "Fourteenth Five-Year Plan" at present.The improvement of reliability is a core problem to be solved in the research and development of domestic control system.Hardware redundancy is an important guarantee for the reliability of control system,which can realize fault elimination and fault tolerance in function.Its cost is lower than maintenance,and has occupied a dominant position in industrial application.In order to improve the reliability of gas turbine control system and promote the process of localization process of heavy-duty gas turbine,the following researches are carried out in this paper:(1)Based on the analysis of the thermal dynamic cycle of gas turbine operation,the functions and main roles of the control system are studied.The types of failure of the system are sorted out,while the detailed models and basic functions of each component of the control system are studied.At the same time,reliability evaluation indexes are analyzed,and the most appropriate research methods and analysis tools for reliability research are selected.(2)The fault tree model of domestic control system is established.By corresponding the components of the fault tree to the logical structure of the system,the fault tree model of the whole control system is established.The effect of redundancy measures such as hot standby and voting for fault tolerance purposes on reducing the failure rate of the system is analyzed.The simulation model is constructed based on the fault tree simulation software Item QRAS,which can analyze the optimization effect of redundancy structure in the system quantitatively.(3)The bayesian belief network model of the domestic control system is established.The events of the fault tree are mapped to the nodes of the bayesian belief network model,and the logical relationships of the control system are mapped to the conditional probability table of the bayesian belief network to obtain the bayesian belief network model.Based on the simulation software GeNIe,the posterior probabilities of the bayesian belief network model are obtained,which not only verifies the accuracy of the fault tree analysis results,but also finds the weak links by ranking the posterior probabilities in order to shorten the maintenance time and improve the system reliability.(4)The dynamic reliability model of domestic control system is established.Further considering the maintenance rate of the system,the original static reliability model is optimized into a dynamic reliability model.which is named markov model.Based on the simulation software probabilistic model checker PRISM,the markov model of the system is analyzed and examined quantitatively to obtain the availability of the system,and the simulation experiment shows that it meets the requirement of the reliability design of the system.