A Maintenance-prediction Method for Aircraft Engines using Generative Adversarial Networks

Due to high monitoring costs and restrictions to the monitoring environment, little condition monitoring data exists that can inform maintenance predictions for aircraft engines. As obtaining sufficient monitoring data is important for optimizing maintenance times, this paper proposes a method for using generative adversarial networks (GANs) to generate condition monitoring data of aircraft engines. The GANs method is introduced and an algorithm for generating monitoring data is designed. The proposed method is used to extend a sample of aircraft engine condition monitoring data. A proportional hazards model with a Weibull distribution and reliability analysis is used to analyze multivariate state parameters and accurately predict the engine's optimal maintenance time. The results of applying the method ultimately confirm its effectiveness.