基于数据挖掘的电站运行参数目标值优化  

The Summary of Optimal Operation Parameters in Power Station Based on the Data Mining

作　　者：王秋平[1] 陈志强[1] 魏浩[1] 

机构地区：[1]东北电力大学自动化学院,吉林吉林132012

出　　处：《电力科学与工程》2015年第7期19-24,共6页Electric Power Science and Engineering

摘　　要：为提高电站经济性和机组运行效率,降低机组发电煤耗,求取电站机组运行参数最优值是关键技术。以往通过理论计算得到最优运行参数值是在设定的理想环境下得到的,在实际的电站运行过程中难以实现。而数据挖掘算法是从电站自身的历史数据中得到的最优运行参数值,电站机组能够很容易在实际运行中实现该值。通过对比近年来电站常用数据挖掘算法,总结出基于数据挖掘的电站优化运行的主要步骤为关联规则、数据离散化、运行工况划分、粗糙集知识约减。得出以下结论:模糊关联规则挖掘算法是电站数据挖掘中的最主要方法,能够适用于大多数的电站优化目标值挖掘;模糊聚类离散化能够克服边界划分过硬的问题,将电站中的连接参数离散化;粗糙集属性约减能够有效降低数据挖掘的参数维度,提高挖掘效率。同时指出基于数据挖掘的电站优化运行算法将成为电站运行参数优化的主要研究方向。In order to increase the economy and the operation efficiency of the power plant and decrease the coal consumption of electrical facilities, it is of great significance to get the optimal operation parameter value of power station. The previous optimal operating parameters were obtained by theortical calculation under the ideal environment, which were difficult to achieve in the actual operation of the power plant. However, the data mining algorithm is a way to get the optimal operating parameters from previous data, which can easily obtained in the actual operation. By comparing previous data mining algorithm of power plant in recent years, the paper summarized the main steps of optimal parameters by data mining, which include association rules, data diseretization, condition i- dentification and knowledge reduction. In addition, it concluded that fuzzy association data mining is the main method of data mining in power plant, which can be applied to power stations for optimization value mining and fuzzy clustering discretization can disperse the continuous data of parameters in power plant. Rough set theory can reduce the dimension of parameters and improve the efficiency of data mining. Finally, the result shows the optimization of parameter based on data mining algorithm may provide guidenee for optimal running future research.

关 键 词：关联规则 数据离散化 工况划分 知识约简 

分 类 号：TP274.2[自动化与计算机技术—检测技术与自动化装置]