科研之友
微信
新浪微博
Facebook
分享链接
摘要
Considering spatial-temporal characteristics of alarm information in power systems, this paper presents a novel fault diagnosis model based on DM(data mining) of sequence of events(SOE) to study fault diagnosis of high-voltage transmission line system(HVTLS). The proposed model makes full use of the relevance among spatial-temporal characteristics of SOE information. Dynamic programming algorithm is adopted for editing operation to measure the similarity between event sequences. Then fault diagnosis problem is converted into the seeking of the most similar normal event sequence to real time fault event sequence with the lowest operation cost. The simulation shows that distorted fault patterns can be restored to normal fault pattern, and the higher fault-tolerant performance of fault diagnosis system can be ensured in the presented approach.
