The Intelligent Monitoring Key Technologies Developed by Dadu River Company Reached International Advanced Level

On December 5, the outcomes of the “knowledge graph-based signal intelligent monitory key technologies” research project of Dadu River Company, a subsidiary of CHN Energy, passed the assessment and inspection of the China Society for Hydropower Engineering. Upon comprehensive evaluation, experts of the assessment team agreed that the research project explored new models for hydropower operation monitoring and accelerated scientific and technological progress in the hydropower industry, with its outcomes reaching the international advanced level.

Targeting at the hydropower signal monitoring scenario, the research project carried out in-depth exploration on knowledge graph representation, construction and application methods in the field of hydropower operation. For the first time in China, it put forward the grading and classification solutions for the intelligent monitoring technology of hydropower generation and operation, and built knowledge graphs on hydropower generation and operation signal monitoring. With Dadu River Company’s Pubugou hydropower station as a pilot project, it formed a graph pattern with equipment, signals, states, and movements as the core factors. Moreover, based on the realities of production scheduling, the research project focused on hydropower entity extraction models and relation extraction methods, creatively developed supervision and learning-based methods for automatic extraction of knowledge about hydropower generation and operation, and achieved preliminary automatic acquisition of triplet knowledge about hydropower generation and operation. At the same time, it developed the real-time signal analysis and reasoning model for the hydropower monitoring system, established the signal analysis, judgment and processing mechanism, and basically formed an intelligent signal monitoring system with functions such as signal relationship recognition, feature detection, and noise signal perception. This is of great significance for strengthening the safe and efficient operation of cascade hydropower clusters and shows great potential for widespread utilization.