摘要
城市照明监控历史运行数据往往蕴含着大量的潜在信息和知识,人们迫切需要对有价值的数据进行深度挖掘,并将获得的成果应用于运行状况评估、异常预警和运营参数调优中。基于城市照明监控历史运行数据,提出了一种基于大数据分析技术的应用方法,对海量运行数据进行聚类分析,以及对场景模式进行划分得到判别决策树,并对实时监测过程中的动态数据进行离群点分析,从而判别当前设备运行状况。结合应用实例对模型进行合理性验证,证明了该方法的可行性。
Urban lighting control historical operating data often contain~ a l^rlle number of potential mtormanon anu knowledge, we urgently need to carry out the depth of mining valuable data, and will obtain results for running status as- sessment, early warning and operation parameters optimization in abnormal. Based on historical operating data of urban lighting control, we propose a method based on the application of large data analysis techniques, running through the mas- sive data clustering analysis, and then divide the scene mode to get identification tree, the last of the real-time dynamic monitoring process outlier data analysis to determine the current status of equipment operation. This paper combines ap- plication example to verify the reasonableness of the model to prove the feasibility of this method.
出处
《软件导刊》
2015年第5期1-4,共4页
Software Guide
关键词
大数据分析
聚类分析
判别决策树
离群点分析
Big Data Analysis
Cluster Analysis
Discriminant Tree
Outlier Analysis
