Anomaly Detection of Aerospace Facilities Using Ganomaly

In the field of aerospace, the abnormal detection of data is of great significance. The rapid and effective detection of abnormal parameters is key to find potential failures of spacecraft. Traditional methods of anomaly detection need much manual labour and material resources but cannot satisfy the requirements of real-time accuracy. At the same time, there are far more normal samples than abnormal samples, and the original classification methods cannot be applied. In this paper, we propose a GANomaly-based framework for anomaly detection of aerospace data. GANomaly is a framework that analyzes the underlying relationships of data using potential space, which is more in line with the characteristics of the payload data and the actual scenarios for anomaly detection. This article compares GANomaly with other anomaly detection methods on the public aerospace dataset and payload dataset respectively. The results show that the GANomaly-based anomaly detection framework has good capabilities for detecting abnormality of aerospace datasets.