【摘要】 锅炉各监控参数基准值的确定是分析锅炉运行能耗偏差的基础。该文充分利用锅炉运行数据的关联特性,提出了一种基于模糊C-均值聚类算法实现多参量同步聚类以确定锅炉监控参数基准值的方法。该方法可以在实际运行数据中同步挖掘出某典型负荷邻域区间对应的排烟氧量、排烟温度和飞灰含碳量等监控参数基准值,从而达到改善锅炉运行性能的目标。在多参量同步聚类算法中,利用有效性函数优化模糊聚类数,提出运行模式支持度的相关概念及其样本支持判定的规则,并对类中心点处较小ε区域内样本进行无偏估计。实例分析结果表明:该方法能够在兼顾参数之间耦合关系的基础上,得到高效工况下对应的各基准值样本点,并建立相应的基准值模型。更多还原

【Abstract】 Determining target-value of boiler monitoring parameters is groundwork of analyzing operational energy consumption deviation.Considering the relevance among operating data,a method of determining target-value of boiler monitoring parameters based on fuzzy C-means(FCM) clustering algorithm was proposed in this paper.From practical operating data,target-value of several monitoring parameters such as exhaust-gas oxygen,unburned carbon content and exhaust-gas temperature in some typical load neighborhood regions were excavated synchronously by this method,it’s conducive to improvement of boiler operation.In the process of data mining by this method,the number of fuzzy clusters was optimized by validity function,concept about operational mode support and its regulation of judging sample support were presented,and then the samples in ? region of cluster centre were unbiased estimated.Results of case study showed that this method can determine target-value in high efficiency operation mode at the base of considering coupled relation among the parameters,and its models can be also set up.更多还原