基于LSSVM-MODE的锅炉效率与NO_x排放优化研究  

STUDY ON OPTIMIZATION OF BOILER EFFICIENCY AND NO_x EMISSSON BASED ON LSSVM-MODE

作　　者：刘定平[1] 蔡泓铭[1] 叶向荣[1] 

机构地区：[1]华南理工大学电力学院,广东广州510640

出　　处：《热力发电》2010年第7期23-26,35,共5页Thermal Power Generation

摘　　要：超临界机组锅炉效率与NOx排放之间存在着互相制约的关系。通过建立最小二乘支持向量机(LSSVM)模型对某超临界600 MW机组锅炉运行数据进行训练,并通过多目标微分进化算法(MODE)对锅炉效率及NOx排放进行寻优,得到了全局优化后参数调整策略。优化结果表明,此模型能很好地兼顾降低NOx排放量和提高锅炉效率的目的,为超临界火力发电机组高效低污染运行调整提供指导。There are mutual restrictions between boiler efficiency and NOx emission for supercritical boilers.Through establishment of least square support vector machine（LSSVM） model to train the operation data of boilers for supercritial 600 MW unit,and through the method of multi-objective differential evolution（MODE） algorithm,the boiler efficiency and NOx emission have been optimized,and the adjustment strategy after overall optimization being obtained.Results of optimization show that the said model can take both boiler efficiency and NOx emission into consideration,providing guidance for highly effective and low NOx emission in operation adjustment of supercritical thermal power units.

关 键 词：支持向量机 超临界 600MW机组 锅炉效率 NOX排放 多目标优化 

分 类 号：TK224.11[动力工程及工程热物理—动力机械及工程]