Abstract
Conference Title: 2017 IEEE 7th Annual International Conference on CYBER Technology in Automation, Control, and Intelligent Systems (CYBER) Conference Start Date: 2017, July 31 Conference End Date: 2017, Aug. 4 Conference Location: Honolulu, HI, USA Based on the research and construction of smart grid at home and abroad, this paper puts forward the application of machine learning theory to smart grid. For the first time, this paper tries to integrate the traditional power grid strategy with the computer network to establish the distribution network fault solving strategy library, leading the grid to be intelligent. At the same time, another innovation of this paper is to put forward the conception of the twin model system. The method abstracts the power grid into the integration of the information-physical system through data exchange, and the grid state in each node corresponds to the twin matrix, which makes the grid modeled, digitized. According to this idea, the smart grid system can find the optimal strategy in the fault solving strategy library after the failure, and can realize the self - healing function of the distribution network. In this paper, we first simply introduce the algorithms of machine learning and information physics coupling system. Then we introduce the composition of twin model system and its basic principles. Finally, we elaborate the process of intelligent distribution network's offline construction and online operation.