Performance parameter prediction technology is the core research content of aeroengine health management,and more and more machine learning algorithms have been applied in the field.Regularized extreme learning machin...Performance parameter prediction technology is the core research content of aeroengine health management,and more and more machine learning algorithms have been applied in the field.Regularized extreme learning machine(RELM)is one of them.However,the regularization parameter determination of RELM consumes computational resources,which makes it unsuitable in the field of aeroengine performance parameter prediction with a large amount of data.This paper uses the forward and backward segmentation(FBS)algorithms to improve the RELM performance,and introduces an adaptive step size determination method and an improved solution mechanism to obtain a new machine learning algorithm.While maintaining good generalization,the new algorithm is not sensitive to regularization parameters,which greatly saves computing resources.The experimental results on the public data sets prove the above conclusions.Finally,the new algorithm is applied to the prediction of aero-engine performance parameters,and the excellent prediction performance is achieved.展开更多
It is desired to increase the core engine speed of the turbofan, to get the best efficiency for the next leap of the engine technology. The conventional mechanism in which the front fan is directly connected to the ou...It is desired to increase the core engine speed of the turbofan, to get the best efficiency for the next leap of the engine technology. The conventional mechanism in which the front fan is directly connected to the output shaft of the core engine has a limit of increasing the spool speed because the fan diameter is very large. The authors have proposed a new driving system in which the front fan is driven through the aerodynamic torque converter. The front fan can work at the conventional speed while the core engine runs more efficiently at higher speed. Continuously, in this paper, the flow through the converter is simulated numerically by CFX-5 with the k-εturbulence model of the commercial CFD code. The secondary flow occurred on the hub wall affects markedly the flow condition on the blade surfaces, and the flow along the suction surface of the driver blade separates near the trailing edge, which is deviated to the blade tip by the centrifugal force due to the wheel rotation.展开更多
文摘Performance parameter prediction technology is the core research content of aeroengine health management,and more and more machine learning algorithms have been applied in the field.Regularized extreme learning machine(RELM)is one of them.However,the regularization parameter determination of RELM consumes computational resources,which makes it unsuitable in the field of aeroengine performance parameter prediction with a large amount of data.This paper uses the forward and backward segmentation(FBS)algorithms to improve the RELM performance,and introduces an adaptive step size determination method and an improved solution mechanism to obtain a new machine learning algorithm.While maintaining good generalization,the new algorithm is not sensitive to regularization parameters,which greatly saves computing resources.The experimental results on the public data sets prove the above conclusions.Finally,the new algorithm is applied to the prediction of aero-engine performance parameters,and the excellent prediction performance is achieved.
文摘It is desired to increase the core engine speed of the turbofan, to get the best efficiency for the next leap of the engine technology. The conventional mechanism in which the front fan is directly connected to the output shaft of the core engine has a limit of increasing the spool speed because the fan diameter is very large. The authors have proposed a new driving system in which the front fan is driven through the aerodynamic torque converter. The front fan can work at the conventional speed while the core engine runs more efficiently at higher speed. Continuously, in this paper, the flow through the converter is simulated numerically by CFX-5 with the k-εturbulence model of the commercial CFD code. The secondary flow occurred on the hub wall affects markedly the flow condition on the blade surfaces, and the flow along the suction surface of the driver blade separates near the trailing edge, which is deviated to the blade tip by the centrifugal force due to the wheel rotation.
