In real life, most data sets are unbalanced data sets, while the performance of traditional sampling classification algorithms is poor. In order to improve the detection rate of minority samples, a multi classification fusion algorithm of unbalanced data sets based on Gaussian Mixture Model (GMM) is proposed. Firstly, Gaussian Mixture Model clustering is used to generate minority samples. Then, fo...Show More
Centralized feeder automation has high flexibility and adaptability and is suitable for various distribution system. In this paper, the fault search table is established based on fault information, then we use Aprion data mining technology to analyze the fault search table to locate the fault. This location method can be applied to centralized feeder automation. At the same time, considering the m...Show More
In the existing substations, when the fault occurs, too many alarm signals will be sent to the substation layer, so that the monitoring personnel cannot find the critical fault information in time. In order to solve this problem, this paper proposes a fault diagnosis method using analyzed alarm signals in substation layer. Firstly, use the string matching algorithm to extract the information of al...Show More
