考虑到基于智能计算的燃烧优化算法难以应用于工程领域的问题，设计了一种基于数据挖掘案例推理的电站锅炉燃烧优化系统。根据海量的分散控制系统（Distributed Control System,DCS）历史数据，采用改进的模糊减法聚类算法确定分类数，以模糊C均值算法建立初始案例库，通过目标寻优约简案例库。在线应用时，基于非可控因子的案例推理方法计算当前工况最佳燃烧参数，根据可控因子对输出参数进行修正，保证系统的实时最优性。对比某机组的优化效果表明，优化后选择性催化还原（Selective Catalytic Reduction,SCR）装置入口NO〖HT5”〗x浓度平均降低26.2 mg/m3，锅炉效率平均提升了0.21%。

Since the intelligent optimizationbased combustion optimization algorithm is difficult to be applied in engineering field,this paper designs a boiler combustion optimization system based on data mining casebased reasoning.Based on the massive DCS historical data,this system uses the improved fuzzy subtraction clustering algorithm to determine the classification number,and establishes the initial case library by the fuzzy Cmeans algorithm.Finally,it seeks to optimize the case base through the target.Then,the casebased reasoning method based on noncontrollable factors is used to calculate the optimal combustion parameters of the current working conditions,and the outputs are corrected according to the controllable factors to ensure the realtime optimality of the system.Compared with the optimization effect of a certain unit,NO〖HT5〗x concentration at SCR inlet is reduced by 26.2 mg/nm3,and boiler efficiency is increased by 021%.

TK221

内蒙古电力（集团）有限责任公司科技项目(2018-73)