Based on big data, machine learning, knowledge graphs and other technologies, the system provides a closed-loop end-to-end data and AI solution that allows electric power companies to analyze data, get an insight into data and make better decisions. At the bottom layer is the data middle platform that offers metadata management capabilities to enable access to multiple-source heterogeneous data, quality management, standardization, and data development and sharing; what is at the middle layer includes the knowledge graph and AI platforms. The knowledge graph platform provides services throughout the knowledge lifecycle from model building, production, management, and inference; the AI platform provides a one-stop model development platform for developers by leveraging NLP, CV, and ML technologies. With data accumulation, knowledge acquisition, and model development as the cornerstone, intelligent applications including Power Transformation Signal Fault Identification, Workload Capacity-Based Scheduling, Intelligent Writing, and Electric Power Expert are available at the application layer.
