Liquid rocket engine thrust chamber fault diagnosis method based on augmented particle filtering

Abstract

The invention discloses a liquid rocket engine thrust chamber fault diagnosis method based on augmented particle filtering, and the method comprises the steps: considering a common fault effect according to the working principle of a thrust chamber, and building a thrust chamber nonlinear thermodynamic nominal model and a fault model; designing a particle filter according to the nonlinear thermodynamic nominal model, and estimating the working state and output of the thrust chamber; constructing a residual error between estimation output and measurement output of the thrust chamber, and achieving fault detection of the thrust chamber by adopting a generalized likelihood ratio test method; integrating the fault estimation vector into the state estimation vector, expanding the dimension of the state estimation vector, and establishing an augmented fault identification particle filter to realize fault identification; and according to a fault identification result, designing a fault classification algorithm, establishing an improved augmented fault identification particle filter by using a classification result, and improving the fault estimation precision. According to the invention, rapid and high-precision fault detection and diagnosis can be carried out in real time under external interference, random system noise and observation noise.

Description

technical field [0001] The invention belongs to the technical field of liquid rocket engine fault diagnosis, and in particular relates to a liquid rocket engine thrust chamber fault diagnosis method based on augmented particle filtering. Background technique [0002] As the most important part of the propulsion system of the launch vehicle, the rocket engine's safe and reliable operation is the key to the success of the launch mission. However, its complex working environment and harsh working conditions make it one of the most fault-prone parts in the entire space transportation system. Among them, because the thrust chamber is in a high-temperature and high-pressure environment and directly communicates with the outside atmosphere, it has the characteristics of large external interference, strong random noise, and fast fault development. According to the statistics of liquid rocket engine failures in recent years, the thrust chamber is second only to the turbo pump, and i...

AI Technical Summary

Problems solved by technology

[0003]The model-based Kalman filter method requires the system to be a linear Gaussian noise system, and the data-based wavelet analysis method requires enough samples, especially the number of fault samples, AI-based methods are less interpretable

Therefore, none of the above three methods can complete the task of thrust chamber fault detection and diagnosis well.

Images