Technologies Used by General Electric to Create Digital Twins for Energy Industry

The paper discusses the aspects of implementation of high-precision digital twins for power plants and energy systems based on the approaches used by General Electric (GE), which has an extensive experience in this field. To create such twins, the company uses laws of physics, knowledge and experience in engineering design, methods of measuring and controlling various parameters, as well as artificial intelligence technologies. The paper offers a description of models, methods, and approaches used by GE when developing digital twins. The company’s experience in the use of artificial intelligence technologies, such as pattern recognition, unstructured (multimodal) data analysis, and expert networks, as well as sensor technologies is presented. The most important component of GE’s digital twin is the Predix simulation platform. It was specifically developed by GE by using industrial Internet technologies to perform analytical processing of huge amounts of machine data in real time. The basic capabilities and specifications of the Predix platform are provided.