On September 14, Longyuan Power Engineering Technology Company, relying on the scientific and technological innovation project of "National Energy Group's new Energy Investment and construction integration data analysis research and Application Demonstration", completed the large-scale deployment of distributed intelligent algorithm platform nationwide for the first time in the industry, and improved the intelligent monitoring means of safe production, operation and management of Longyuan Power.

The company has built an open, intensive, cloud-edge full-stack intelligent algorithm platform in the way of "one center, multiple nodes". The system takes the actual production and operation scenario of Longyuan Power as the entry point, combines computer vision perception to conduct all-round supervision and management of personnel production operations, equipment operating status and environmental unsafe factors, and continuously empowers safe production through intelligent application. In addition, the system combines production work order to achieve optimal scheduling, automatic triggering of algorithm model, automatic call of equipment video, and automatic development of intelligent analysis, comprehensively promoting business upgrade and ensuring maximum utilization of server resources. Through the mode of "Model deployment - early warning analysis - alarm aggregation - model training - model delivery - model deployment", the closed-loop management of the model is realized, the continuous optimization and self-learning ability of the algorithm model are realized, and the accuracy of the algorithm model is improved.

The distributed intelligent algorithm platform realizes the organic combination of unstructured data such as video and audio and safe production, further digs the potential value of production and operation data, builds an open ecology of new energy intelligent application, and lays a solid foundation for promoting the intelligence and digitalization of new energy plants and stations. (Yang Wei)

Source: China Electric Power News