An LSTM Neural Network-Based Fault Line Detection System Using CloudPSS-XStudio

This paper designs a fault line detection system deployed on a digital twin application platform, CloudPSS-XStudio. which can easily deploy distributed neural network models on web software to meet the requirement of application platform engineering landing. First, original multi-source fault record data or simulation data is organized and stored in the CloudPSS data platform as a basic training database. Then, the proposed LSTM neural network-based fault line detection kernel is created and integrated into CloudPSS FuncStudio. Finally, a web-based application is built on AppStudio, which integrates the original data and analysis kernel to provide an online fault line detection service. Test results demonstrate the effectiveness of the designed platform. The performance is tested using benchmark test data with high accuracy.