An Equipment Condition Warning Method Based on MEP in Smart Substation

Jiansheng Li, Zhicheng Zhou, Yiming Wu, Yuncai Lu, Chao Wei, Peng Wu
2015 Proceedings of the 2015 International conference on Applied Science and Engineering Innovation   unpublished
To monitor the equipment condition changes, a warning method based on MEP (Maximum Entropy Principle) is proposed. First, the data is normalized and the data objects are divided into K clusters by K-means clustering algorithm. Second, the parameters are calculated and its probability density function is obtained by maximum entropy principle. Last, the new parameter data is compared with warning data to decide the condition changes. In this method, clustering operation can reduce the impacts of
more » ... uce the impacts of outside environments and probability density function based on maximum entropy principle avoids the short comings of subjective assumption. Monte Carlo simulations verify the effectiveness and accuracy.
doi:10.2991/asei-15.2015.411 fatcat:liy4mca5ovcr5p5mr2gs6dffre