Fault Diagnosis For Start-up Procedure Of Gas Turbine

Gas turbine is currently sharing more application in the fields such as aviation and navigation. Currently, electricity generating with gas turbine is also becoming an inevitable tendency. As gas turbine works under condition of high temperature, high pressure and high rotor speed, fault is prone to happen in the device. When it comes to start-up process, which bears large impact of high roter variation and high temperature variation, the fault is more possible to happen. Currently, it is trend to apply fault diagnosis technology on gas turbine to ensure safety in operation. So it is obvious that enhancing fault diagnosis of gas turbine start-up process has important significance.Diagnosis of device fault relies on large number of operation data. However, for start-up process, which duration is short compared with a steady state operation, it is impossible to acquire large number, multi kinds of fault data. If simulation can be applied to the process of data collection, it is possible to analyze start-up fault of different kinds or ranges.In this paper, a mathematical model which can simulate start-up procedure of gas turbine is built on the knowledge of gas turbine theory. Fault of different kinds and ranges have been set in the model, and typical start-up fault such as thermal blockage, compressor surging and ignition failure are simulated. Different fault degrees have been set, and indexes such as surge margin, rotor speed, accelerated speed, start-up duration and fuel air ratio have been calculated, with which the influence of fault degree on start-up process can be evaluated. Furthermore, fault diagnosis based on time series fragment extraction is implemented, the results impresses the efficiency of this diagnosis algorithm. In the end, three kinds of start-up control methods are designed to prevent thermal blockage caused by degration of fuel low heat value, and all of these methods help to prevent the fault.