Ding Jun
2 Papers
Ding Jun is an academic researcher. The author has contributed to research in topics: Transformer & Voltage. The author has co-authored 2 publications.
Chat about Author
Papers
Patent
Electronic transformer error risk assessment method and device based on regression neural network
Chen Gang,Lu Shufeng,Li Zhixin,Yang Shihai,Xu Minrui,Ding Jun,Chen Wenguang,Zhou Gan,Li Zhi,Cheng Siyuan,Lu Zigang,Cheng Guofeng,Wu Qiao,Xu Xin,Cheng Hanmiao,You Wenzheng,Chen Fei +16 more
- 31 Dec 2019
TL;DR: In this article, an electronic transformer error risk assessment method and device based on a regression neural network is presented. But the method is characterized by comprising the steps of training an electronic transform risk assessment model based on the regression neural networks according to the current and voltage of an electronic transformer and screened environmental characteristic data; inputting the characteristic data of the electronic transformer to be assessed into the model to obtain the predicted ratio difference and angle difference of the transformer, and calculating the operation risk index of the transformer according to predicted ratio differences and angle differences.
Patent
Electronic transformer credibility evaluation method and device based on whole-network domain evidence set
Huang Qifeng,Lu Shufeng,Xu Minrui,Li Zhixin,Yang Shihai,Chen Wenguang,Chen Gang,Cheng Guofeng,Zhou Gan,Liu Xi'ang,Ding Jun,Xu Xin,Lu Zigang,Cheng Siyuan,Wu Qiao,Cheng Hanmiao,Yao Gandong +16 more
- 31 Dec 2019
TL;DR: In this article, an electronic transformer credibility probability evaluation method and device based on a whole-network domain evidence set is presented. But the method comprises the steps of clustering high-dimensional data sets of an EH transformer to obtain different clusters of the clustered high dimensional data sets, and removing noise point data; outputting a predicted credibility probability according to the credibility probabilities of the highdimensional data set of the electronic transformer, and training a pre-established XGBoost model to obtain a trained XGBOost model.