电网总调调度信息披露不平衡性数据处理模型  

Data processing model for unbalanced data disclosure of power grid overall dispatching information

作　　者：侯方迪 高卫东 张勇[1] 翟哲 杨凡 吴亮 HOU Fangdi;GAO Weidong;ZHANG Yong;ZHAI Zhe;YANG Fan;WU Liang(China Southern Power Grid Co.,Ltd.,Guangzhou 610106,China)

机构地区：[1]中国南方电网有限责任公司,广东广州610106

出　　处：《电子设计工程》2022年第17期104-108,共5页Electronic Design Engineering

摘　　要：针对电网总调调度数据信息量剧增,导致信息披露系统的数据信息处理性能较差的问题,构建了一种电网总调调度信息披露不平衡性数据处理模型。通过数据清洗、数据标准化以及数据平衡化对不平衡性数据进行预处理,利用基于人工蜂群算法提取数据特征,构建决策树分类器,并组合成随机森林模型,实现不平衡性数据处理。以红水河水库日来水数据为例,对数据进行不平衡数据处理,结果表明,与对比方法相比,应用该文构建的模型,水库日来水数据的不平衡数据处理结果更好,得到的水库日来水数据趋势更为准确,并能够有效预测未来的日来水情况。Aiming at the problem that the data and information processing performance of the information disclosure system is poor due to the sharp increase in the amount of data and information of power grid dispatching,a data processing model of unbalanced information disclosure of power grid dispatching is constructed.The unbalanced data is preprocessed through data cleaning,data standardization and data balancing,and the data features are extracted based on artificial bee colony algorithm,and the decision tree classifier is constructed,which is combined into a random forest model to realize the unbalanced data processing.Taking the daily inflow data of Hongshuihe reservoir as an example,the unbalanced data processing is carried out on the data.The results show that compared with the comparison method,the unbalanced data processing result of the daily inflow data of the reservoir under the application of the model constructed in this paper is better,the trend of the daily inflow data of the reservoir is more accurate,and it can effectively predict the future daily water situation.

关 键 词：电网总调调度信息 不平衡性数据 处理模型 决策树分类器 

分 类 号：TP253.5[自动化与计算机技术—检测技术与自动化装置]