Fault Diagnosis based on Behavior and State-Awareness Digital Twin of Key Equipment in Power Supply System

With the rapid development of multi-electric and all-electric aircraft, the role of power supply systems in aircraft is becoming increasingly prominent. Traditional fault diagnosis methods have problems such as a single means of sensory modelling and unbalanced fault data. The rapid development of digital twin technology provides an opportunity to overcome these difficulties. However, how to achieve adaptive updating and how to improve the data-and-model-fusion capabilities are also urgent challenges to be solved. To address the lack of existing research, this paper combines time-frequency domain analysis, physical information neural network based on differential-algebraic equation (DAE-PINN) and transient stability analysis of the power supply system based on a digital twin model with dimensions including fault-behavior-twin and (FBT) fault-state-awareness-twin (FSAT). Ultimately, this paper achieves effective digital twin modelling, fault eigenfeature extraction and high accuracy fault diagnosis, reaching fault data balance and complete state characterization and adaptive update of the diagnostic model.