数据挖掘与锅炉负荷多模型自适应控制研究  

Data Mining and Multiple Model Adaptive Control of Boiler Load

作　　者：卢勇[1] 徐向东[1] 

机构地区：[1]清华大学

出　　处：《电站系统工程》2003年第4期49-51,共3页Power System Engineering

摘　　要：针对目前锅炉负荷控制中的难点问题,提出了利用多模型自适应控制策略加现代数据挖掘技术的方法加以解决:前者本着“分而制之”的思想,通过建立覆盖锅炉各种运行工况和操作条件的模型集和选取适当的切换规则,完成各种条件下的负荷自适应控制;后者从锅炉对象实际运行数据入手,利用数据挖掘技术,提取有关锅炉运行状态、操作条件与燃料发热量之间的关联规则,形成相应的专家知识库用于在线获取燃料发热量信息,优化模型集,指导模型切换。仿真试验表明该方法相对于传统方法具有更佳的控制性能。Aiming at the difficulty of boiler load control problem, a novel method combining the multiple-model adaptive control (MMAC) strategy with data mining technique is presented. Through construction of the model set covering all kinds of working and operation condition, MMAC strategy based on the 慸ivide and conquer?principle, can tackle the complicated boiler load control problem; from the practical boiler running database, some rules which associate the boiler working and operation condition with the fuel heat value can be extracted through data mining technique, and can be utilized to optimize the model set. The simulation result indicates that this method is superior to the conventional one.

关 键 词：数据挖掘 锅炉 负荷控制 自适应控制 关联规则 热力发电厂 

分 类 号：TM621.2[电气工程—电力系统及自动化] TP273.2[自动化与计算机技术—检测技术与自动化装置]