In response to the current problem of dispersed power marketing data and lack of unified management of power marketing data within the power marketing business system,a multi-source log security data mining method for power marketing was adopted,and Apache Lucene's Elasticsearch distributed search engine was applied in the multi-source power marketing database.By applying the Cortex-A9 processor with the main control chip model XC7Z035FFGH676-2,the mining and computing capabilities of multi-source power marketing security data information in power marketing were improved;the clustering analysis method of self-organizing mapping neural network and fuzzy clustering algorithm improved the ability to detect anomalies in power marketing data;the use of self-organizing mapping neural networks and fuzzy clustering algorithms reduces energy data consumption and improves data mining capabilities.The clustering analysis time of the proposed method in this paper is as short as 104 seconds,laying the foundation for the next step of research.
关键词
电力营销/聚类分析/模糊聚类算法/神经网络/自组织映射/异常检测
Key words
power marketing/cluster analysis/fuzzy clustering algorithm/neural network/self-organizing map/anomaly detection
维普
万方数据