基于数据融合的烟气含氧量软测量技术综述

A Review of the Soft Sensing Technology for the Flue Gas Oxygen Level Based on Data Fusion

软测量技术在过程控制领域中有重要的应用,对于关键的且难于检测的过程变量,研究如何对其建立软测量模型有着重要的意义。将数据融合技术应用到软测量的数据处理上,有效提高了软测量预测值可靠性和准确性。锅炉烟气含氧量是计算锅炉过量空气系数的关键参数。在实际生产过程中,用于测量此参数的氧化锆传感器可靠性不佳。针对这一问题,关于烟气含氧量的软测量建模研究硕果连连。在参考大量相关文献的基础上,对其研究现状及发展前景做了简单的归纳。然而糖厂锅炉有其自身的特殊性,其燃料有蔗渣和煤两种,而蔗渣燃烧时,过量空气系数通过CO2的检测更为简单,这就对软测量建模提出了新的要求。

The soft - sensing technology has a very important application in the field of the process control. It is very significant to study how to establish the soft sensing model for the process variables which are important but difficult to test. With the data fusion methods applied to the soft - sensing technology, the soft model＇ s forecast value are getting more and more accurate and reliable. The boiler flue gas oxygen level is a key parameter during the process of calculating boiler ex- cess air coefficient. In the actual operation process, the zirconium oxide sensors which are used to detect the flue gas oxy- gen levels have very poor reliability. In order to solve this problem, many researchers begin to study how to establish the soft models of flue gas oxygen levels, and have got many achievements. On the basis of many relevant literatures, it has a brief summary of the research status and development prospect for the above study. Guangxi sugar mills, however, has their own particularity. There are two kinds of fuels ： bagasse and coal. It is much more simpler to calculate the excess air coeffi- cient by the flue gas level of C02when bagasse burned. This has put forward the new request for the models of soft sensing.