基于分层分块状态相关独立成分分析的火力发电过程运行状态评价

Operation performance assessment of thermal power generation process based on HMPRICA

火力发电承担着我国电能供应的重要任务,实时掌握火力发电过程运行状态,对于提高煤炭资源利用率,实现综合经济效益的提升具有重要意义.针对火力发电过程,提出一种新的基于分层分块状态相关独立成分分析的运行状态评价方法.该方法将火力发电全流程进行纵向和横向的层次结构划分.在利用分层分块状态相关独立成分分析建立评价模型时,不仅能够有效提取子系统内部与过程运行状态密切相关的特征信息,还能够提取不同子系统之间的相关关系和相互作用对全流程运行状态的影响.所提出方法能够降低评价模型的复杂度,改善评价结果的可解释性.当过程运行状态非优时,根据子系统的评价结果和变量贡献率精准定位非优原因.最后,以某火力发电过程为背景验证所提出方法的有效性.

Thermal power generation process undertakes the important task of power supply.Grasping the operating performance of thermal power generation process has great significance for improving the utilization rate of coal resources and the comprehensive economic benefits.A new assessment method based on hierarchical multiblock performancerelevant independent component analysis(HMPRICA)is proposed in this study.The whole process is divided into multilayer and multiblock in the horizontal and vertical directions.The assessment models based on HMPRICA can not only effectively extract the variation information closely related to the process operating performance,but also extract the influence of the interaction among different subsystems.The proposed method can reduce the complexity of each assessment model and improve the interpretability of the assessment results.When the process is nonoptimal,the responsible causes can be accurately located according to the subsystem assessment results and variable contribution rate.Finally,an actual thermal power generation process is taken as the background to verify the effectiveness of the proposed method.