发电锅炉飞灰含碳量软测量建模及燃烧优化运行研究

发布时间：2019-03-15 21:03

【摘要】：随着工业生产和社会经济的飞速发展,节能环保问题日益受到人类的重视。火力发电的燃煤锅炉在燃烧过程中不仅要消耗大量的能源,还会排出大量的废气和有害烟尘,对大气环境造成严重的污染。因此,实现能源的高效利用是火力发电行业实现节能减排的一项有效措施。飞灰含碳量是衡量火力发电锅炉燃烧效率高低的一项重要指标,目前,国内电力生产企业大部分都采用人工取样、制样和实验室化学分析的方法进行离线检测,而对于锅炉燃烧优化工程上通常也只凭借操作经验进行燃料和配风的人工调整,难以达到理想的运行效果,导致了能源的浪费,因此开展此项课题的研究具有重要的工程意义。本文以发电燃煤锅炉为对象,分析了锅炉生产过程的工艺特点,介绍了飞灰含碳量监测和锅炉燃烧优化的研究现状,并深入分析了影响飞灰含碳量相关因素,在此基础上,采用改进的BP神经网络算法建立了飞灰含碳量预测模型,并依据一种新的群体智能算法───狼群算法对锅炉燃烧工况进行优化,最后通过仿真和实测数据验证了预测模型和燃烧优化方法的有效性,具有较高的工程应用价值。本文主要研究内容如下:1.通过对国内外研究现状的调研综述,分析了发电燃煤锅炉燃烧过程的工艺特点,并归纳了有关飞灰含碳量影响因素以及降低飞灰含碳量常用方法。2.针对BP神经网络预测飞灰含碳量存在的样本误差问题,对神经网络误差函数进行改进设计,并验证了对于输入样本中的干扰具有较好的抑制作用。3.采用主元分析方法对神经网络模型进行精简,针对飞灰含碳量测量输入变量过多的问题,分析了各输入变量对输出变量的贡献值,筛选了网络的输入参数;设计了基于主元分析的BP神经网络飞灰含碳量预测模型,并进行了仿真分析和实验验证。4.提出了基于狼群算法的锅炉燃烧优化方案。依据飞灰含碳量的预测数据,利用狼群算法对燃烧工况进行优化,选择最有利于燃烧的控制方案,并对其进行了仿真研究。
[Abstract]:With the rapid development of industrial production and social economy, the problem of energy saving and environmental protection has been paid more and more attention by human beings. Coal-fired boilers with thermal power not only consume a lot of energy, but also emit a large amount of exhaust gas and harmful smoke and dust, which cause serious pollution to the atmospheric environment. Therefore, the realization of efficient use of energy is an effective measure to achieve energy conservation and emission reduction in thermal power industry. The carbon content of fly ash is an important index to measure the combustion efficiency of thermal power boiler. At present, most domestic power production enterprises use manual sampling, sample making and laboratory chemical analysis to carry out off-line detection. For boiler combustion optimization engineering, it is difficult to achieve ideal operation effect by manual adjustment of fuel and air distribution only by virtue of operation experience, which leads to waste of energy. Therefore, it is of great engineering significance to carry out the research on this subject. Taking coal-fired boiler for power generation as an object, this paper analyzes the technological characteristics of boiler production process, introduces the research status of carbon content monitoring of fly ash and optimization of boiler combustion, and deeply analyzes the related factors affecting carbon content of fly ash, on the basis of which, the paper introduces the research status of carbon content monitoring of fly ash and optimization of boiler combustion. Based on the improved BP neural network algorithm, the prediction model of carbon content in fly ash is established, and the combustion condition of boiler is optimized according to a new swarm intelligent algorithm called wolf swarm algorithm. Finally, the effectiveness of the prediction model and combustion optimization method is verified by simulation and measured data, which has high engineering application value. The main contents of this paper are as follows: 1. Based on the survey of the current research situation at home and abroad, the process characteristics of combustion process in power generation coal-fired boilers are analyzed, and the factors affecting the carbon content of fly ash and the commonly used methods to reduce the carbon content of fly ash are summarized. 2. Aiming at the problem of sample error in prediction of carbon content in fly ash by BP neural network, the error function of neural network is improved, and it is verified that the neural network error function has a better inhibitory effect on the interference in input samples. 3. The neural network model is simplified by principal component analysis. Aiming at the problem that there are too many input variables in carbon measurement of fly ash, the contribution value of each input variable to the output variable is analyzed, and the input parameters of the network are selected. The prediction model of carbon content in fly ash based on BP neural network based on principal component analysis is designed and simulated and verified by experiments. 4. A boiler combustion optimization scheme based on wolf swarm algorithm is proposed. According to the prediction data of carbon content in fly ash, the combustion condition is optimized by using wolf swarm algorithm, and the most favorable control scheme is selected, and the simulation study is carried out.
【学位授予单位】：安徽工业大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TM621.2

【参考文献】