基于聚类算法的双进双出磨煤机入炉煤量软测量方法

Soft measurement method of coal quantity for double-inlet and double-outlet mill based on clustering algorithm

双进双出磨煤机制粉系统由于结构复杂,制粉过程惯性大,锅炉入炉煤量难以准确测量,严重影响机组自动控制系统调节性能。为此,本文提出一种双进双出磨煤机入炉煤量软测量方法。该方法在双进双出磨煤机制粉系统动态特性基础上,利用磨煤机煤粉流量与容量风门开度、一次风压力、料位等关键参数的相关性,采用分段线性函数构造了锅炉入炉煤量软测量模型,采用聚类算法通过机组历史运行数据辨识该软测量模型相关特性函数的参数,并在某330 MW机组控制系统优化项目中成功实现了入炉煤量的实时测量。该方法能实现入炉煤量的准确测量,有效提高机组自动控制系统性能。

Because of the complex structure and the large inertia of pulverizing process,the feed coal quantity of the boiler is difficult to be accurately measured,which seriously affects the regulating performance of automatic control system of the unit.Therefore,a soft measurement method of feed coal quantity is proposed for the boiler with double-inlet and double-outlet mills.Based on dynamic characteristics of the coal pulverizing system,this method establishes a soft measurement model of coal quantity of the boiler with the help of piecewise linear function by using the correlation between the pulverized coal flow rate and the key parameters such as the opening of the capacity damper,the primary air pressure and the material level,and uses the clustering algorithm to identify the relevant characteristic parameters of the soft measurement model through historical operation data of the unit.In the control system optimization project of a 330 MW unit,the real-time measurement of the feed coal quantity of the boiler has been successfully realized.This method can realize accurate measurement of feed coal quantity and effectively improve the performance of the automatic control system of the unit.