Computer Engineering and Applications ›› 2018, Vol. 54 ›› Issue (5): 265-270.DOI: 10.3778/j.issn.1002-8331.1609-0025

Previous Articles    

Real-time fault monitoring of transmission lines based on Spark

CHEN Jianxia1, ZHU Jiqi1, ZHANG Yue1, ZHANG Xiaoxing2, LV Juntao3, BAI Demeng3   

  1. 1.School of Computer Science, Hubei University of Technology, Wuhan 430068, China
    2.School of Electrical Engineering, Wuhan University, Wuhan 430072, China
    3.State Grid Shandong Electric Power Research Institute, Jinan 250002, China
  • Online:2018-03-01 Published:2018-03-13


陈建峡1,朱季骐1,张  月1,张晓星2,吕俊涛3,白德盟3   

  1. 1.湖北工业大学 计算机学院,武汉 430068
    2.武汉大学 电气工程学院,武汉 430072
    3.国网山东省电力公司 电力科学研究院,济南 250002

Abstract: Since the monitoring data of transmission lines are the largest part of the amount of data in the smart grid, including not only the online condition monitoring data, but also the basic information of the devices, the experimental data, defect records, it requires a higher performance of the reliability and real-time in the data processing. The paper designs and realizes a novel model to solve the real-time fault monitoring of transmission lines according to the practical application of power transmission line faults’ types. In particular, the paper constructs a distributed cluster based on Spark, an efficient real-time data processing system, for the transmission line fault real-time monitoring, develops a distributed ISODATA and fuzzy KNN big data analysis algorithm. Compared with standalone KNN algorithm, it improves 70.75% efficiency of the time performance. Experimental results demonstrate the proposed approach has the obvious advantages of the computational efficiency.

Key words: real-time big data, transmission lines, fault monitoring, distributed Iterative Self Organizing Data Analysis Techniques Algorithm(ISODATA) algorithm, distributed fuzzy k-Nearest Neighbor(KNN) algorithm

摘要: 输变电线路状态监测数据是智能电网中数据量很大的一部分,不仅包括在线的状态监测数据,还包括设备的基本信息、实验数据、缺陷记录等,在数据处理的可靠性和实时性方面的要求都很高。根据实际应用中输变电线路的故障类型,设计并实现了输变电线路实时数据故障监测模型。其中,利用高效处理实时数据的Spark系统,研发出基于Spark的分布式ISODATA和模糊KNN大数据分析算法,与单机KNN算法相比,在时间性能上提高了70.75%效率,具有明显的计算效率优势。

关键词: 实时大数据, 输变电线路, 故障监测, 分布式迭代自组织数据分析算法(ISODATA), 分布式模糊k最近邻分类算法(KNN)