Power equipment status information parallel fault diagnosis of based on MapReduce

摘要

A diagnosis algorithm against status information parallel faults occured on power equipment is analyzed and explored based on MapReduce. Here comes the parallelized Naive Bayes algorithm and transformer DGA test which help understand this algorithm by 4 MapReduce processes, coupled with the dissolved gas parameters of transformers given here. It turns out that the transformers can be diagnosed by Hadoop clusters when they reach a certain scale, as shown in the experiment.