The Early Warning For The Fault Of High Temperature Components In Gas Turbine

As the first of the national "13th Five-Year" development plan 100 tasks, gas turbine will be in a surprising state of development, while receiving extensive concern about the security and stability of the running because of working conditions and all kinds of failures. It is most representative that high temperature gas turbine components. The high temperature gas turbine components are made of Flame tube, transition section, guide bushing, interconnector, fuel nozzle, spark plugs, flame detector and turbine blades. It will cause great influence to the management production that the failures of high temperature gas turbine components caused by working for a long time in high temperature and h igh pressure environment. Therefore, early warning for high temperature gas turbine parts plays a very important significance.The relationship between gas turbine exhaust temperature and high temperature parts, the key factors influencing the exhaust temp erature, the establishment of early warning model, the analysis of the change of high temperature parts performance are studied in the thesis. The main work and results are listed as follows:(1) Due to the high combustion engine temperature, temperature s ensors cannot directly measure temperature. The exhaust temperature of turbine is usually used to measure temperature of high temperature parts in an indirect way. But because of large amounts of information contained in exhaust temperature, indirect measurement method cannot assure enough precision. The correlation analysis and gas turbine running mechanism were used to analysis the effect of abnormal high temperature parts to exhaust temperature. For instance, fuel nozzle clogging will result in changes of gas turbine exhaust temperature distribution. Then the main parameters that result in the change of discharge temperature are determined, which there is a strong linear relationship between measuring point discharge temperature and average temperature. As a conclusion, if the obvious linear relationship changes, abnormal conditions will happen in the high temperature parts.(2) Based on support vector regression method, the high temperature combustion engine parts early warning model is established. Based on the multi parameter combinatorial optimization of genetic algorithm, the precision of the early warning model is improved. Then based on the actual gas turbine data, the selection rules of the model’s training set are identified, that the data, a wide range of changes in working conditions of gas turbine during a short time, is used to build model so that the model will have the good prediction accuracy. Based on the confidence interval and the improved quality control chart, early warning mechanism for high temperature combustion engine parts is determined. The alarm system can not only guarantee the robustness of the model and reduce the false alarm rate of the model.(3) In addition to early warning for high temperature combustion engine parts, computer network and communication technology are utilized to build online monitoring early warning system for the high temperature parts, based on which 8 sets of gas turbine have been monitored for the real-time. As a result, the system has the feasibility and stability.