基于Pearson系数与多元线性拟合的330MW汽轮机高压汽缸轴向膨胀定量监测方法

High-pressure Cylinder of 330MW Turbine Based on Pearson's Coefficient and Multivariate Linear Fitting Quantitative Monitoring Method for Axial Expansion

高压汽缸轴向膨胀反映了汽轮机缸体在运行过程中的受热状态,是表征机组发生异常动静碰摩和滑销系统卡涩潜在风险的重要参数,也是机组运行期间重点监测的对象之一。然而,实际中不少服役时间较长的机组缺少可全面反映轴向缸胀的测点,现有研究中也缺乏对轴向缸胀的定量分析。针对一台330MW汽轮发电机组的实际情况,根据轴向缸胀对滑销系统的作用机理,设计了增加测点方案并进行工程实施,获取了多端轴向缸胀变化的测量数据。在此基础上,提出了基于Pearson相关性系数和多元函数线性拟合的方法,建立汽轮机左右两端轴向缸胀与各高相关性测点的方程。定量分析结果显示:函数拟合优度都高于0.998。最后,设计了一套轴向缸胀在线监测系统,可实现关键参数的实时综合监测,对实际故障预警具有重要的价值和意义。

The axial expansion of the high-pressure cylinder reflects the heating state of the turbine cylinder during operation,which is an important parameter to characterize the potential risk of abnormal dynamic and static friction and jamming of the slip-pin system in the unit,and it is also one of the key objects to be monitored during the operation of the unit.However,in practice,many units with long service time lack measurement points that can comprehensively reflect the axial cylinder expansion,and the quantitative analysis of axial cylinder expansion is also lacking in the existing studies.In this paper,for the actual situation of a 330MW turbine-generator unit,according to the mechanism of axial cylinder expansion on the slip-pin system,we designed a program to increase the number of measurement points and carried out the engineering implementation,and obtained the measurement data of the change of axial cylinder expansion at multiple ends.On this basis,a method based on Pearson's correlation coefficient and linear fitting of multivariate function is proposed to establish the equations of axial cylinder expansion at the left and right ends of the turbine with each high correlation measurement point.The quantitative analysis results show that the goodness of fit of the functions are all more than 0.998,respectively.Finally,a set of online monitoring system for axial cylinder expansion is designed,which can realize the real-time comprehensive monitoring of the key parameters,and is of great value and significance for the actual fault early warning.