【摘要】 大型火力发电厂的实时控制过程中,许多数据之间呈现复杂的非线性关系,而数据挖掘技术能从数据中发现知识或规则,及时分析、调整参数。该文利用数据挖掘方法确定监控参数的基准值,为火电机组耗差分析提供重要依据。介绍了数据挖掘方法的相关理论,研究并应用聚类算法确定热力设备监控参数的基准值模型。采用k-means法分析实时运行数据,分别挖掘再热器压损和锅炉排烟温度的基准值,所得锅炉运行主要监控参数的基准值模型经样本图证实效果良好。为基准值模型的确定提供了一个新的思路和有效方法。更多还原

【Abstract】 Complicated nonlinear relationships exist among many data in the real-time control-process of large power plant.And data-mining technology could find knowledge,analyze parameters and regulate them.This paper ascertained target-value by means of data-mining,which supported energy-loss analysis.The paper introduced relative theory on data mining,studied and applied target-value model of thermal supervised parameters in the way of clustering algorithm.k-means arithmetic was utilized to analyze real-time operating data of thermal units,and mined target values of pressure loss in reheater and exhaust-gas temperature from boiler.The target-value models for main supervised parameters of boiler were received,which was proved to be effective.And the results supply a new idea and effective method for target-value models.更多还原