基于全工况数据挖掘的多目标燃烧优化  

Multi-objective combustion optimization based on data mining with full-scale working condition

作　　者：郑伟[1] 刘达 ZHENG Wei LIU Da(School of Mechatronical Engineering and Automation, Tianjin Vocational Institute, Tianjin 300410, China Tianjin Guodian Jirmeng Co-generation Co., Ltd., Tianjin 300300, China)

机构地区：[1]天津职业大学机电工程与自动化学院,天津300410 [2]天津国电津能热电有限公司,天津300300

出　　处：《热力发电》2017年第6期69-74,共6页Thermal Power Generation

摘　　要：为了使燃煤电站锅炉燃烧达到高效低污染的目的,提出了锅炉燃烧优化的约束模糊加权规则提取算法,该算法可以快速确定燃烧过程中各变量的定量关系,并给出相应的数据挖掘方法,从锅炉全工况海量历史运行数据中提取燃烧优化规则,并构建燃烧优化知识库,实现多目标燃烧优化。通过在发电机组上的实际实验,验证了在同等工况下,利用锅炉燃烧优化规则,NOx排放质量浓度平均值下降69.47 mg/m3,飞灰含碳量平均值下降0.05%,锅炉效率平均值提升0.11%。因此,基于全工况数据挖掘的多目标燃烧优化,能够有效缓解燃煤电站节能减排的压力,也便于实施和在线应用。To realize high combustion efficiency and low pollution emission for coal-fired utility boilers, the constrained fuzzy weighted extracting rules algorithm was proposed for boiler combustion optimization, which can quickly determine the quantitative relationship among different combustion variables. Moreover, with the algorithm at the core, a data mining approach was presented to extract the optimal rules of boiler combustion from massive historical operation data covering full-scale working condition. A knowledge base for multi-objective combustion optimization could be developed by all the optimal rules. Experiments on the generator units proved that, by using this boiler combustion optimization rule, the NOx emission decreased by 69.47 mg/m3 on average, the carbon content in fly ash reduced by 0.05% on average and the boiler efficiency increased by 0.11% on average, comparing with the operation results without optimal rules. Therefore, the data mining approach can be effective in relieving pressure on energy conservation and pollution reduction and is convenient to be implemented in practice.

关 键 词：燃煤电站 燃烧优化 数据挖掘 NOX排放 飞灰含碳量 锅炉效率 

分 类 号：TM611[电气工程—电力系统及自动化]