基于高阶特征提取的气路故障诊断方法OA

Gas turbines are characterized by complex structures,highly coupled components and long-term operation in high-temperature and high-pressure environments,making it difficult for sensors to directly detect the specific location of faults.To address the challenge of diagnosing gas path faults in gas turbines,this study proposes a method based on operational big data from power plants.The method involves modifying characteristic curves of compressor and turbine components,extracting high-order features from these curves and incorporating the historical operating conditions of the components to assess their performance,thereby achieving gas path fault diagnosis.Furthermore,to improve diagnose efficiency and reduce labor costs,an intelligent platform is independently developed,and an online performance monitoring model is established.This model enables real-time,rapid and accurate calculation of performance indicators,allowing direct comparison and judgment.The proposed fault diagnosis method is applied to a domestic F-class gas turbine unit,and the application results have verified the reliability and advancement of the method,demonstrating its effectiveness in diagnosing gas path faults.The research findings can provide technical reference for optimization and upgrading of gas path fault diagnosis technologies and construction of intelligent monitoring platforms for similar F-class gas turbines.They can also offer theoretical support and practical guidance for operation safety and stability,fault diagnosis,efficient operation and maintenance of gas turbine units in power plants.