A construction method of a knowledge graph of power system based on various data-driven missions

Inhalt:
We propose a method of construction of a knowledge graph (KG) of power system based on various data-driven missions. In this knowledge graph, devices and their attributes, topology information and electrical and meteorological measurements are included. Based on the KG, we can first summarize and classify the missions in the system operation. Besides, machine-learning (ML)-model information is also involved, and KG will recommend suitable ML methods for new tasks. Therefore, KG assists to improve the generalization ability of machine-learning models among various conditions and different missions. Simulations on a modified IEEE-14 bus system demonstrate the efficacy of the knowledge graph dealing with different tasks in the power system.