基于生产数据挖掘的吹灰需求度置信规则库研究  

Confidence rule base for soot blowing demand based on production data mining

作　　者：钱虹[1] 宋亮[1] 陈琪琪[1] 陈纲[2] QIAN Hong SONG Liang CHEN Qiqi CHEN Gang(College of Automation Engineering Shanghai University of Electric Power, Shanghai 200090, China Huaneng Shanghai Shidongkou First Power Plant, Shanghai 200942, China)

机构地区：[1]上海电力学院自动化工程学院,上海200090 [2]华能上海石洞口第一电厂,上海200942

出　　处：《热力发电》2017年第6期113-118,共6页Thermal Power Generation

基　　金：上海市自然科学基金(15ZR1417500)~~

摘　　要：为对锅炉积灰结渣受热面吹灰优化,以锅炉燃烧特性和吸热特性为理论依据对生产数据进行挖掘处理,建立吹灰需求度置信规则库。运用数理统计、贝叶斯理论,结合现场运行专家经验,完成吹灰策略集和征兆集的提取、规则的表示和规则变量的设定,并对规则库进行测试。结果表明,本吹灰需求度置信规则库对领域专家吹灰经验的知识描述符合电厂实际吹灰需求,达到了吹灰优化的目的。Aiming at optimizing the soot blowing for boiler, a confidence rule base was established by data mining, on the basis of combustion characteristics and heat absorption characteristics of the boiler. By using mathematical statistics and the Bayesian theory algorithm combined with the experts＇ operation experiences, the soot blowing tactic and symptom set was extracted, the rules are expressed and the rule variables were set. Moreover, the rule base was tested. The results show that, this confidence rule base meets the actual demand of soot blowing with description of the experts＇ experience and knowledge representation, which basically realizes soot blowing optimization.

关 键 词：锅炉 积灰结渣 吹灰优化 置信规则库 数据挖掘 贝叶斯理论 生产数据 

分 类 号：TK16[动力工程及工程热物理—热能工程]