Fault Diagnosis Method for Mobile Energy Storage Cabin Based on Digital Twin Technology and Deep Autoencoder

A fault diagnosis method for mobile energy storage cabin based on digital twin technology and deep autoencoder is proposed to address the problems of timeconsuming, labor-intensive, and low accuracy in traditional fault diagnosis methods. Firstly, a fault diagnosis model based on digital twin technology is constructed based on the mechanism and data model of mobile energy storage cabin, and the reliability of diagnosis is improved through the combination of virtual and real. Then, a deep autoencoder model is designed, which adopts an adaptive data filtering mechanism based on an improved spectral clustering algorithm, and a fusion loss function is designed to optimize the model parameters. Finally, the deep autoencoder model is deployed in the digital twin model to achieve fault diagnosis by analyzing energy storage cabin data. Based on the selected mobile energy storage cabin data, experimental analysis is conducted on the proposed method, and the results shows that the average accuracy and time consumption of fault diagnosis are 96.38% and 3.73s, respectively, demonstrating significant advantages in state diagnosis.