The artificial intelligence in the most thermal po

AI "landing" in power is in the Ascendant: it can be used for power dispatching

original title: AI "landing" in power is in the Ascendant: it can be used for power dispatching

on the morning of the 18th, "2018 Zijin power discussion - International Symposium on intelligent power protection and operation control" was held in Nanjing. Chinese and foreign power experts attended the meeting. Photo by Zhu Xiaoying

China news, Nanjing, August 18 (Zhu Xiaoying) the application of artificial intelligence in CEC has just started. According to an interview with a number of power experts on the 18th, the current application of artificial intelligence in CEC has covered the fields of power load forecasting, electric hidden danger fault forecasting in extreme weather, photovoltaic hardware "health" status forecasting, large power dispatching simulation and so on

on the morning of the 18th, "2018 Zijin on electricity - International Symposium on smart electricity protection and operation control" was held in Nanjing. The theme of this symposium is "intelligent power protection and operation control", which is co sponsored by the National Electric Power Research Institute Co., Ltd. and the State Key Laboratory of intelligent power protection and operation control. Power experts from China, the United States, Russia, the United Kingdom, Germany, Japan, Canada, New Zealand, Brazil and other countries attended the meeting

Shanxin from the system research and development center of NARI Group Research Institute is an expert who "lands" artificial intelligence in the electrical system. He told that the application of artificial intelligence in China's power industry has just started and is still being explored

on the morning of the 18th, "2018 Zijin on electricity - International Symposium on smart electricity protection and operation control" was held in Nanjing. Chinese and foreign power experts attended the meeting. Zhu Xiaoying photographed

for example, they once "fed" Ai different "raw materials" such as line trip data, typhoon path, tower wind resistance parameters caused by typhoons in East China in recent years, so that it could learn and compare. As a result, they found that compared with traditional algorithms and physical modeling methods, the accuracy of AI "predicting" the probability of line trip caused by typhoon was improved by 20%

Shan Xin said that at present, the more mature application of AI is in the field of operation inspection. For example, the inspection robot in Jiangsu can inspect the abnormal power lines. Future robots that can complete experiments on pipe ring stiffness, ring flexibility, flattening, weld stretching and so on are expected to break through the existing "active boundaries" and complete tasks in deep mountains, tunnels and underground

for the future, he believes that artificial intelligence is applicable to the scenario of power dispatching. In the past, the operation mode of thermal power was fixed. With the explosive growth of new energy and distributed photovoltaic projects, the operation of electricity is affected by many uncertain factors. Artificial intelligence has great potential here: through continuous learning and discovery of laws, data-driven improvement of traditional physical modeling methods, and prediction of electrical operation status. These are being realized step by step in China's laboratories

the electric dispatching work also includes a lot of operation experience, "artificial intelligence can independently learn rules, and in the future, it may liberate manpower to realize a new electric dispatching mode like 'autonomous driving'. Figure 4 thin-walled design solution." Shan Xin said

in his view, with the continuous upgrading of hardware, data and algorithms, the application prospect of artificial intelligence in electricity is expected. (end)