故障诊断算法在MapReduce中的优化实现

Optimization and Implementation of Fault Diagnosis Algorithm in MapReduce

针对故障诊断算法特点,给出了MapReduce框架实现故障诊断算法的基本方法;通过对故障诊断算法和MapReduce运算框架的分析,得出诊断算法MapReduce化的基本方法和思路;在算法研究的基础上,针对非迭代诊断算法,采用传统MapReduce框架实现;针对迭代诊断算法,采用添加了传送模块的迭代式MapReduce框架实现;分别以最近邻法和模糊C均值聚类算法为例,给出非迭代和迭代诊断算法MapReduce化的具体实现;实验结果表明,文章所提基本方法可以运用在故障诊断算法中,为诊断算法MapReduce化提供依据,将MapReduce运用到故障诊断算法中可以有效提升故障诊断效率。

According to the characteristics of fault diagnosis algorithm, the basic method of fault diagnosis algorithm based on MapReduce framework is presented. Through the analysis of fault diagnosis algorithm and MapReduce framework, the basic methods are obtained. Used traditional MapReduce framework to achieve non--iterative diagnosis algorithm. Used iteration MapReduce framework to achieve iterarive diagnosis algorithm. Taking the Nearest Neighbor and Fuzzy c--Means Clustering algorithm as example, respectively. The specific implementation based on MapReduce is given. The experimental results show that the proposed method can be used in diagnostic algorithms, and provides a basis for implementing diagnosis algorithm based on MapReduce. Using MapReduce in fault diagnosis algorithm can effectively improve the efficiency of fault diagnosis.