火电机组飞灰含碳量软测量模型研究-控制科学与工程；控制理论与控制工程专业论文.docxVIP

摘 摘 要 摘 要 随着世界的持续发展，能源短缺日益突出，能源的开发和有效利用程度是生 产技术和生活水平的重要标志，也是我国经济发展的一个重要的问题。目前，电 力行业已经成为我国最为重要的能源产业，因此，努力提高发电效率和降低发电 成本成为我国各大发电企业发展方向的重要任务。 锅炉飞灰含碳量是反映火力发电厂燃煤锅炉燃烧效率的一项重要指标，精确 和实时地监测飞灰含碳量有利于提高锅炉燃烧控制水平，降低发电成本，提高机 组运行的经济性，同时也有利于提高煤灰的品质，促进煤灰的商品化。但是由于 工业水平有限，传统的离线测量技术具有样品不具有代表性、运行不稳定和测量 数据实时性低等缺陷，不能有效的指导锅炉运行，从而不能很好的提高锅炉效率。 本文提出一种基于BP神经网络建模的飞灰含碳量软测量技术，并加入基于机理 建模分析实时在线监测的煤质参数与煤元素的辅助变量，使本文模型不同于往常 的仅仅基于运行数据的神经网络模型，并在最后利用主元分析对输入变量进行优 化处理，对输入进行降维，降低模型的复杂性，使模型更加精确和可靠。实验结 果表明，基于BP神经网络的本文优化模型与传统模型相比，具有很好的泛化能 力，计算速度快且准确，是飞灰含碳量的很好的一种软测量模型。 关键词：火电机组；煤质参数；飞灰含碳量；软测量；主元分析 万方数据 摘 摘 要 Abstract With the continued development of the world，energy shortages have become increasingly．prominent，and the effective utilization of energy is all important symbol of the development of production technology and living standards，is also an important issue of economic development．At present，。the power industry has become China’S most important energy industry,therefore，strive to improve the power genermion efficiency and reduce the cost of electricity has become an important task of the development direction of the maj or power companies in China． Unburned carbon。in fly ash is an important indicator reflecting the boiler combustion efficiency of coal—fired power plant，accurate and real-time monitoring of unburned carbon will help improve boiler combustion control levels．reduce electricity costs and improve the economics of plant operation，but also help to improve the taste of ash，soot promote commercialization．However,due to limited industrial level，the tradkional off-line measurement technology has sample is not representative，unstable operation and measurement data real-time and low defect，can not effectively guide the boiler operation，which Can not be good to improve the efficiency of the boiler．in this paper,a soft measurement technique of unburned carbon in fly ash is proposed based BP neural network modeling，and adding auxiliary variables of coal quality parameters and elements of real—time online