Research on acoustic emission diagnosis of diesel combustion chamber components
Based on acoustic emission signal and transfer learning,a new fault diagnosis method for diesel combustion chamber is presented in this paper.In this paper,the oil injector blockage,valve pressure reduction and exhaust valve leak-age fault are simulated on TBD234V6 diesel engine.CompactRIO hardware is used to collect signal,and the characteristics of AE signal after the failure of combustion chamber components are analyzed.Research shows that fault diagnosis method based on feature parameter extraction and transfer learning can identify different fault types more accurately.Compared with traditional machine learning algorithm,fault diagnosis method has higher accuracy,stronger generalization ability,and bet-ter adaptability to the case of fewer data samples and different data distribution.This study is of great significance to ensure the health of diesel combustion chamber components and safe navigation of ships.
NETL
NSTL
万方数据
