Gas turbine analytic redundancy construction method based on variable weight neural network

Abstract

The invention provides a gas turbine analytic redundancy construction method based on a variable weight neural network, and belongs to the field of intelligent aero-engine control, and the method comprises the steps: collecting the operation process data of a gas turbine; making an input data graph; constructing a variable weight neural network; training the network by using the data graph; and based on the trained variable weight neural network, real-time accurate estimation of the analysis redundancy is realized. For multiple variables and strong nonlinearity of the gas turbine, the analytical redundancy construction method provided by the invention has the advantages of strong nonlinear expression capability and high calculation speed, and is suitable for fault diagnosis and fault-tolerant control of the gas turbine.

Description

technical field [0001] The invention proposes a gas turbine analytical redundancy construction method based on a variable weight neural network, which belongs to the field of intelligent aeroengine control. Background technique [0002] Aeroengine control system sensor fault diagnosis and fault-tolerant control usually use hardware redundancy, but with the increase of hardware redundancy, the weight and volume of the engine will inevitably increase, and even reduce the overall performance of the system. In order to avoid the above problems, the analytical margin is generally used as the margin information, that is, according to the internal relationship between the various variables of the engine, the measured value of a certain parameter can be estimated by the rest of the sensors, and the estimated value is the analytical margin, so It can provide a reference value for sensor fault diagnosis, and judge whether the fault occurs according to the residual error between the es...

AI Technical Summary

Problems solved by technology

For tasks of the same difficulty, because the variable weight neural network has a stronger nonlinear expression ability, it requires fewer operational parameters and a faster calculation speed, which can overcome the problems of slow calculation speed and poor real-time performance of deep learning algorithms; The variable weight neural network model is used to estimate the analytical margin, which can solve the problem of slow calculation of the analytical margin of the gas turbine

Images