【作者】 易正明；

【导师】 周孑民；

【作者基本信息】 中南大学 ， 热能工程， 2007， 博士

【摘要】 熟料在回转窑中的煅烧是烧结法生产氧化铝工艺中的关键工序，对氧化铝的产量和质量有着直接的影响。近年来，我国氧化铝工业产量不断增加，对产品质量的要求也日益提高。因此，深入了解熟料烧结过程中的各个影响因素，实现回转窑生产过程的自动控制对提高氧化铝产量与质量，降低工人劳动强度，实现生产节能降耗，促进氧化铝工业的技术进步具有重要的理论与实践意义。由于回转窑生产过程具有大惯性、大时滞、非线性、时变、多干扰等特征，且窑体处于旋转状态，温度在线测量困难，参数检测严重滞后。目前氧化铝回转窑生产过程控制中所应用的集散控制系统仅能起到监测以及一些常规控制作用，没有充分利用信息资源，这些问题影响了回转窑生产过程优化与智能控制的实现。在现有工艺流程和生产设备的情况下，在线仿真、图像处理、模糊控制、软测量等先进过程控制技术对解决上述问题具有独特的优越性。本文在对回转窑热平衡和传热过程进行分析的基础上，通过对窑筒体非稳态传热过程以及窑内物料流动和传热过程的研究，证明了窑筒体传热过程在一个蓄热、放热周期内可以简化为稳态换热以及窑内同一横截面上物料内部温度分布基本趋于一致，为建立回转窑一维传热数学模型提供了理论依据。通过研究导出了回转窑内物料自然堆角的计算公式，建立了物料与被覆盖窑壁间热传导的物理数学模型，并求出了物料与被覆盖窑壁间的热传导综合传热系数，推导了窑内各辐射换热量的计算公式和变导热系数多层筒壁的温度表达式，并结合煤粉燃烧、物料热解、窑壁散热和窑内传热传质方程，最终建立了回转窑传热综合数学模型。结合回转窑实际生产过程，开发了回转窑尾部烟气流量检测仪和窑中弧形滑环测温仪，实现了回转窑尾部烟气流量和窑中温度的在线检测，并针对回转窑窑内衬和窑皮形状复杂的特点，采用红外线测温仪进行筒体表面温度测量，结合数字滤波补偿技术实现了回转窑状态监测，有效地改善了回转窑的操作，提高了设备运转率。以“全息仿真”理论为基础，以现场在线检测参数为边界或初始条件，通过综合运用炉窑热工及反应动力学原理，首次开发了回转窑在线仿真优化模型及软件。该系统能实时计算回转窑各带长度及窑内温度分布，并由计算机在线显示，记录运行历史曲线，为现场的考核和管理提供依据。在此基础上，结合回转窑工艺的特点与工程技术人员、熟练操作工人的经验，建立了回转窑状态预报系统，实现了回转窑的数据输入、自动报警和故障显示等功能，为生产运行和操作提供了在线指导。基于图像的连通性和相关性原理，结合邻域平均法提出了一种改进的梯度倒数加权平滑算法，更好的实现了图像的预处理，并深入比较和研究了C均值、模糊C均值和快速模糊C均值聚类方法在回转窑火焰图像分割中的应用。在预处理滤波的基础上，首次利用快速模糊C均值聚类方法对回转窑图像进行了准确的分割，实时提取了物料高度、火焰温度和物料温度等特征值，以量化的形式较全面反映了窑内的热工信息，为回转窑的操作优化和自动控制提供了可靠的依据。通过完善相关硬件和通讯等基础设施，首次将回转窑在线仿真、状态预报、图像处理与优化控制结合起来，开发了氧化铝回转窑模糊控制系统。该系统以回转窑火焰温度、物料高度和窑尾温度作为控制的输入量，以燃煤量和排烟机风门开度作为控制的输出量，首次实现了回转窑下煤量和进风量的统一调节，创造了良好的经济效益。氧化铝回转窑热工分析与智能控制系统投运一年多以来，较好地实现了工艺参数的优化以及自动控制，大大减轻了工人的劳动强度，提高了设备运转率，降低了生产成本。根据现场生产数据统计，实现回转窑自动控制后，每吨熟料煤粉单耗节约1.5kg以上，每小时熟料产能提高2t以上，达到了增产降耗的目的。更多还原

【Abstract】 The calcination of stuff in rotary kiln is a key process of aluminaproduction by sintering, which may directly influence the yield andquality of alumina. Along with raised production in our country in recentyears, higher demands are required on alumina quality. For improving thealumina yield and quality, reducing the intensity of labor, decreasingenergy consumption as well as promoting the technical progress ofalumina industry, it is important both in theories and practices to deeplyunderstand influence factors in stuff calcination and to realize automaticcontrol of the rotary kiln production.The process in rotary kiln is characterized with large inertia, longtime delay, nonlinearity, time-dependent and multi-disturbance and thekiln body keeps rotating during the production. It is very difficult toperform temperature on-line measurement, and parameters monitoring isseverely delayed. Furthermore, the distributed control system applied incurrent rotary kiln production of alumina can only fulfill regularmonitoring and some simple control, and the information are not fullyused. All these problems severely harm the process optimization andintelligent control of rotary kiln. But some advanced process controltechnologies such as on-line simulation, image processing, fuzzy control,soft-sensing technique can be used with unique superiority in solving theabove problems.Based on analyzing the heat balance and heat transfer process ofrotary kiln, through studying the transient heat transfer of cylindrical wall,material flow and heat transfer process, this paper proves that the heattransfer process of cylindrical wall can be assumpted as steady state in aenergy storage or energy releasing period, and the temperature of materialon a certain cross section can be treated as the same. It provides thetheory basis for developing an one-dimensional mathematical model forheat transfer of rotary kiln.This research puts forward calculation formula of material naturalangle in rotary kiln, and builds physical-mathematical model of heat conduction between material and covered furnace wall, and gets the heatconduction overall coefficient. It deduces calculation formula of radiationheat transfer and temperature expression of multilayer cylindrical wallwhich possesses temperature dependent thermal conductivity. Combinedwith pulverized coal combustion, material pyrogenation, cooling offurnace wall finally, and heat transfer and mass transfer, this paper buildscombined heat transfer mathematical model of rotary kiln.Combined with the real production process of the rotary kiln, a flowrate meter measuring end gas and an arc cirque temperature measuringinstrument have been developed for realizing the on-line monitoring ofend gas flow rate and temperature in the kiln. According to thecharacteristics of the shape of inner liner and kiln crust, an infrared raytemperature measuring instrument was adopted to collect the temperatureon the surface. A kiln condition monitoring system was developed byintegrating digital filter compensating method. The system improves theoperation of kiln effectively and enhances the efficient of facilities.The on-line simulation and optimization model and software weredeveloped by applying synthetically the principles of kiln thermalengineering and reaction kinetics. This system can realize real-timecalculation of the section lengths and temperature distribution in the kiln,which can be displayed real-time on the computer and recorded ashistorical curves. It offers a tool for in-site examination and management.This information combined with technical characteristics of rotary kilnand experiences of engineer and specialized worker, a rotary kilndiagnosis system was developed for realizing the function of data input,auto alarm and malfunction display etc. This system offers on-lineguidelines for the operation of the rotary kiln.Based on the theory of connectivity and relativity of image,combined with the neighborhood averaging method, this paper presentsan improved gradient inverse weight smoothing algorithm which makesimage pretreatment better. The application of C means, fuzzy C meansand approximate fuzzy C means clustering methods in rotary kiln flameimage dividing was compared and researched throughly. Based on filtering pretreatment, the characteristic values such as the height ofmaterial, the temperature of flame and material can be obtained bydividing rotary kiln image with approximate fuzzy C means clusteringmethod. These values can reflect the thermal condition in the kilnquantitatively and provide reliable data for the operation optimization andautomatic control.By improving basic devices of related hardware and communication,and combining on-line simulation, condition forecasting, imageprocessing and optimization control of rotary kiln, the fuzzy controlsystem of alumina rotary kiln was developed. With the flame temperature,material height, and tail temperature of rotary kiln as input parameters,and the pulverized coal feed and door opening size of smoke exhauster asoutput parameters, the system firstly realized the united adjustment ofpulverized coal quantity and air volume which achieved good economicbenefit.Thermal analysis and intelligent control system for aluminaproduction in rotary kiln has been applied successfully for more than ayear, which realizes technical parameters optimization and automaticcontrol, reduces the intensity of labor, enhances the efficient of facilitiesand achieves the goals of the production increasing and energyconsumption decreasing. According to the in-site statistic data, theconsumption of pulverized coal can save above 1.5 kg per ton stuff andthe production of stuff can improve above 2t per hour after realizing autocontrol of the rotary kiln. The aim to increase production and decreasecost has been achieved.更多还原