Spacecraft intelligent fault diagnosis method based on deep neural network

Abstract

The invention relates to the technical field of spacecraft fault diagnosis, and provides a spacecraft intelligent fault diagnosis method based on a deep neural network in order to realize spacecraft intelligent fault diagnosis, guarantee safe and stable operation of a spacecraft and reduce the detection cost, and the method comprises the steps: firstly, building a fault diagnosis model based on the deep convolutional neural network; extracting fault features from the telemetry data set with strong noise; secondly, pre-training the network by using empirical data of other spacecrafts to obtain initial network parameters; and finally, on the basis of a domain adaptive method in transfer learning, constructing a cost function based on maximum mean difference, and performing parameter readjustment on the network model, thereby improving the accuracy of fault data diagnosis. The method is mainly applied to spacecraft fault detection and diagnosis occasions.

Description

technical field [0001] The invention relates to the technical field of spacecraft fault diagnosis, in particular to the field of spacecraft intelligent fault diagnosis based on a deep neural network. It specifically relates to a spacecraft intelligent fault diagnosis method based on a deep neural network. Background technique [0002] Spacecraft refers to space vehicles that operate in space according to the laws of celestial mechanics and perform exploration and development tasks according to certain requirements, mainly including rockets for launching space vehicles, artificial satellites, space probes, spaceships, space shuttles and various space stations. As our country continues to carry out deep space missions such as lunar exploration programs and Mars exploration, the requirements for the stability, reliability and autonomous operation capabilities of the entire spacecraft system, especially the spacecraft control system, have been significantly increased. However, ...

AI Technical Summary

Problems solved by technology

In terms of spacecraft fault diagnosis technology and deep transfer learning, domestic and foreign scholars have carried out relevant exploration research, but the research on combining the two is still in the preliminary stage, and the current shortcomings are mainly reflected in the following aspects: (1) Spacecraft on-orbit telemetry data has the characteristics of large noise, small samples, and unlabeled data. The cost of data collection and labeling is high. Diagnosis of orbiting spacecraft is a research difficulty

(2) The spacecraft data is difficult to be directly used in the training of the neural network, and it is necessary to consider the appropriate spacecraft data preprocessing method

(3) At present, the fault diagnosis of spacecraft generally adopts the method based on the analytical model, the data utilization is insufficient, and there is a large gap between the space environment simulation and the actual situation, which may lead to inaccurate design parameters and make the diagnosis error

Images