A turbine blade fault testing system and its intelligent fault diagnosis method

Abstract

The invention discloses a turbine blade fault test system and an intelligent fault diagnosis method thereof. The test system includes an air supply system, a turbine main body system, a lubricating system and a measurement control system. When the system is running, the air supply system provides air source for the main turbine system, the lubrication system provides lubricating working fluid for the main turbine system, and the measurement control system is used to control the operation safety of the entire system and measure the required performance parameters. The present invention realizes accurate measurement of aerodynamic parameters and vibration parameters during turbine operation by building a turbine blade fault test system, obtains a large amount of test data of faulty blades, and establishes a fault diagnosis model based on convolutional neural network and deep transfer learning , learn, train and test the aerodynamic and vibration signals of faulty blades, and transfer them to real turbine units, laying the foundation for online fault diagnosis of turbine blades in large generator sets.

Description

technical field [0001] The invention belongs to the field of turbomachinery, in particular to a turbine blade fault testing system and an intelligent fault diagnosis method thereof. Background technique [0002] In large generating sets, turbomachinery plays an important role in the conversion of heat and power. With the development of society, its operating parameters are getting higher and higher. As the "heart" of turbomachinery, blades work in harsh environments and bear the combined effects of centrifugal loads, thermal loads and aerodynamic loads. Once a failure occurs, it will seriously affect the economy and safety of the entire unit. Therefore, it is necessary to carry out real-time online fault monitoring and diagnosis for turbine blades. In actual operation, the monitoring data of turbine units are generally aerodynamic parameters and vibration parameters, so the fault diagnosis of turbine blades is mainly carried out based on the changes of aerodynamic parameter...

AI Technical Summary

Problems solved by technology

The performance of fault diagnosis based on deep learning methods largely depends on the scale and quality of the data set. Conducting tests on real blades will consume a lot of manpower and material resources and is not practical

Therefore, without a large amount of actual operating data, deep learning methods cannot be promoted and applied.

Images