基于历史数据分析的汽轮机调节阀流量特性优化  

Optimization of Flow Characteristics of Steam Turbine Control Valve Based on Historical Data Analysis

作　　者：颜纲要 宫喜鹏 黄焕袍 田彬 马娟 YAN Gangyao;GONG Xipeng;HUANG Huanpao;TIAN Bin;MA Juan(Guoneng Bengbu Power Generation Co.,Ltd.,Bengbu,Anhui 233000,China;CHN Energy ZhiShen Control Technology Co.,Ltd.,Beijing 102211,China;Beijing Engineering Research Center of Power Station Automation,Beijing 102211,China;CHN Energy Suzhou Co-Generation Co.,Ltd.,Suzhou,Anhui 234000,China)

机构地区：[1]国能蚌埠发电有限公司,安徽蚌埠233000 [2]国能智深控制技术有限公司,北京102211 [3]北京市电站自动化工程技术研究中心,北京102211 [4]国能宿州热电有限公司,安徽宿州234000

出　　处：《广东电力》2023年第7期133-138,共6页Guangdong Electric Power

基　　金：国家能源投资集团有限责任公司科技项目(GJNY-21-19)。

摘　　要：汽轮机经过长时间运行或调节阀检修解体后,汽轮机调节阀的实际流量特性曲线将偏离设定值,从而影响机组一次调频和负荷控制能力,严重时可能影响机组的安全性;常用的通过现场流量特性实验进行线性化校正的方式存在精度低、耗时长等缺点。为此,开发数据处理分析算法、稳态工况筛选算法对历史数据进行分析处理,开发分段线性化参数寻优算法以及神经网络算法分别对阀门流量特性非重叠度部分和重叠度部分进行线性化校正。将该方法应用于某600 MW火电机组,结果表明,汽机阀门的流量特性线性度得到了有效的改善,提高了机组的安全稳定性。After the steam turbine has been running for a long time or the regulating valve is overhauled and disassembled,the actual flow characteristic curve of the regulating valve will deviate from the set value,which may affect the primary frequency regulation and load control ability of the unit,even affect the safety of the unit in severe cases.However,there are disadvantages of low precision and long time consumption in commonly used methods for linearization correction through on-site flow characteristic experiments.Therefore,this paper develops the data processing and analysis algorithm and steady-state working condition screening algorithm to analyze and process historical data.Meanwhile,it exploits the piecewise linearization parameter optimization algorithm and the neural network algorithm for linearization calibration of the nonoverlapping part and the overlapping part of the valve flow characteristic.The method is applied to a 600 MW boiler generating unit,and the results show that the linearity of the flow characteristic of the steam turbine valve is effectively improved,and the safety and stability of the unit are improved.

关 键 词：数据挖掘 神经网络 流量特性曲线优化 最小二乘法 数据拟合 

分 类 号：TM621.3[电气工程—电力系统及自动化] TK32[动力工程及工程热物理—热能工程]