电力数据仓库中数据挖掘方法仿真

Simulation of Power Data Warehouse Abnormal Data Mining Method

针对电力数据中包含大差异样本数据集中,数据特征无法构成一致的标准,传统的检测方法对数据进行分析时会发生多中心数据问题,造成重复聚类,使得数据出现错误的分类,无法完成数据的准确检测。为提高检测的准确性,提出了一种蚁群聚类优化方法的电力数据仓库数据检测方法。构建电力数据仓库数据特征预测模型,获取预测数据特征,通过蚁群聚类方法对粗糙数据特征进行分类,去除其中的干扰因素,进而检测出准确的电力数据仓库数据特征。实验结果说明,改进方法可以增强电力数据仓库中数据检测的效率和精度,为电力系统的优化提供了参考。

Research the accurate detection of Power Data Warehouse.The electric anomaly data characteristics in the big differences data set can not consist of consistent standard.The paper put forward a power data warehouse data detection algorithm based on ant clustering optimization.The algorithm was used to construct a features forecasting model of data warehouse and get predict abnormal data characteristics.The coarse abnormal data characteristics were classified using the ant clustering method to detect the accurate data features.The simulation results show that the algorithm enhances power detection efficiency and precision.