基于过程监控的烟气排放软测量预测研究

发布时间：2018-05-25 15:03

本文选题：二氧化硫 + 软测量技术　；参考：《华北电力大学》2015年硕士论文

【摘要】：人类面临着越来越严重的环境问题,大气污染是其中的一个重要方面。火电厂排放的废气和烟尘污染物是大气污染的一个重要来源。为了限制火电厂污染物的排放,国家严格制定了污染物的排放标准。二氧化硫是电厂排放最主要的大气污染物,本课题的研究对象是火电厂石灰石-石膏湿法脱硫系统二氧化硫的排放。本文在对火电厂脱硫系统全面了解的基础上,分析了影响脱硫效率和二氧化硫排放浓度的主要因素。目前,火电厂对烟气污染物的主要监测分析设备是烟气连续排放监测系统,它能将污染物排放的数据及时的反映给电厂和监管部门。由烟气连续排放监测系统得到的污染物排放浓度,是通过专门的污染物分析仪测得的。而本文通过软测量技术,基于脱硫系统的相关运行参数,建立了能够预测脱硫效率和二氧化硫排放浓度的模型。本文的重点是软测量建模阶段,通过对脱硫效率影响因素的分析,并根据本文收集数据的实际状况,选取了浆液pH值、脱硫塔入口二氧化硫浓度、脱硫塔入口烟气温度等八个参数作为软测量建模的输入。选取了BP神经网络和支持向量机两种方法,分别建立了对脱硫效率进行预测的模型。结果表明两种方法都能到达到一定的预测效果,而经参数寻优后的支持向量机模型有着更好的预测性能。支持向量机参数寻优的结果为:惩罚参数取值0.75786,核函数参数取值4.5948。寻优后支持向量机模型预测结果的均方误差和平均相对误差分别为0.179和0.367%。最后,通过OPC(OLE for Process Control)技术,实现了MATLAB与污染源过程监控系统中组态软件的数据交换,使MATLAB和组态王两软件各自的优势得到了充分的发挥。理论上实现了脱硫效率和二氧化硫排放浓度的在线预测。
[Abstract]:People are facing more and more serious environmental problems, and air pollution is one of the important aspects. Exhaust gas and soot pollutants emitted from thermal power plants are an important source of air pollution. In order to limit the emission of pollutants from thermal power plants, the state has strictly formulated emission standards for pollutants. Sulfur dioxide is the main atmospheric pollutant emitted from power plant. The object of this paper is the emission of sulfur dioxide from limestone gypsum wet desulfurization system in thermal power plant. Based on the comprehensive understanding of desulfurization system in thermal power plant, the main factors affecting desulfurization efficiency and sulfur dioxide emission concentration are analyzed in this paper. At present, the main monitoring and analysis equipment for flue gas pollutants in thermal power plants is the flue gas continuous emission monitoring system, which can reflect the pollutant emission data to the power plant and the supervision department in time. The pollutant emission concentration obtained from the flue gas continuous emission monitoring system is measured by a special pollutant analyzer. Based on the operating parameters of the desulphurization system, a model for predicting the desulfurization efficiency and the concentration of sulfur dioxide emission is established by soft sensing technology in this paper. The focus of this paper is the soft sensor modeling stage, through the analysis of the factors affecting the desulfurization efficiency, and according to the actual situation of the data collected in this paper, the slurry pH value and the sulfur dioxide concentration at the inlet of the desulfurization tower are selected. Eight parameters, such as flue gas temperature at the inlet of desulfurizer, are used as input for soft sensor modeling. Two methods, BP neural network and support vector machine, are used to predict desulfurization efficiency. The results show that both methods can achieve a certain prediction effect, and the support vector machine model after parameter optimization has better prediction performance. The results of parameter optimization of support vector machine are as follows: the penalty parameter is 0.75786 and the kernel function parameter is 4.5948. The mean square error and average relative error of the prediction results of the optimized support vector machine model are 0.179 and 0.367 respectively. Finally, through OPC(OLE for Process Control) technology, the data exchange between MATLAB and configuration software in pollution source process monitoring system is realized, which makes the advantages of MATLAB and Kingview software fully play. The on-line prediction of desulphurization efficiency and sulfur dioxide emission concentration is realized theoretically.
【学位授予单位】：华北电力大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：X773

【相似文献】