【作者】 曹琳；

【导师】 曹军；

【作者基本信息】 东北林业大学 ， 林业工程自动化， 2013， 博士

【摘要】 随着世界能源危机的日益严重,国际油价飙升,世界各国寻找石油的替代资源迫在眉睫。我国油页岩资源储量丰富,是石油和天然气的可替代能源之一。因此,利用油页岩制取页岩油再次受到世界各国的广泛关注。油页岩反应炉控制效果的优劣直接影响页岩油的产量,而炉内温度的检测及控制又是油页岩反应炉控制系统中的关键环节之一,因此有必要对油页岩炼油过程中反应炉的温度监控进行深入的研究。本文基于红外热像技术、模型图像处理方法、温度场建模方法,结合油页岩反应炉的工艺特点,对反应炉内的料面温度变化情况进行实时监控,形成一种油页岩反应炉的非接触式温度检测方法。本文对油页岩干馏工艺进行了研究,对目前环形反应炉各种测温方法进行了分析和总结,找出了对环形反应炉进行温度检测所存在的问题和困难。分析和研究了干馏炉内温度的精确控制对油页岩产油率的影响关系,主要分析了干馏炉内温度变化对有机质、页岩油、残炭的影响关系,得到了使页岩油的质量达到最优的温度最佳控制点；分析了干馏炉温度对热解气体的影响关系,找到了使热解气体的产量、质量达到最大值的温度最佳控制点,为进行反应炉不同反应阶段的温度最优控制提供了理论依据。针对反应炉内部环境复杂,反映炉热气流发展状态的料面温度特征难以提取的问题,本文从工艺机理的角度分析了料面红外热图像、十字测温等与料面温度场之间的关系,提出了一种反应炉料面区域温度特征提取方法。并对C-V模型进行了改进,将背景填充技术与改进C-V模型及AOS半隐式方法相结合,提出了基于AOS格式的改进C-V模型及背景填充耦合的红外热图像温度场分割算法,解决了红外热图像温度场的自动分割问题。针对红外热图像各种形状的温度场区域,特别是带有凹形区域的温度场区域,以及热像温度场区域目标和背景颜色相近、边缘不清晰等造成识别困难问题。本文结合了基于边缘的活动轮廓模型和基于区域的活动轮廓模型,提出了一种基于全局最优的活动轮廓模型的红外热像多温度场区域检测方法,实现了对红外热像温度场区域的有效检测及料面中心区域温度特征提取。从工艺机理的角度分析了反应炉红外图像、十字测温等多源信息与料面温度场之间的关系,研究了基于红外热图像的料面温度场等温线等特征提取技术。针对利用单一检测信息难以建立准确的反应炉料面温度场模型的问题,提出了基于信息融合的反应炉料面温度场建模方法,该方法充分利用反应炉十字测温检测信息,以基于两点法的温度动态定标方法对温度场进行定标。仿真结果表明,该方法具有较好的泛化能力和准确度,验证了该方法的可行性和优良性。利用提出的建模方法,建立了基于信息融合的反应炉料面温度场监视系统。系统采用可视化界面显示料面温度场模型,更直观地反映料面温度分布情况,为反应炉操作人员提供了炉况的实时、可靠的参考信息。针对油页岩反应炉可能发生冲顶事故问题,提出了利用红外热成像技术,结合热像不同区域温度变化的趋势及规律,进行了反应炉冲顶事故的预警研究,弄清了反应炉发生冲顶的机理。首先,结合红外热像技术建立了一个反应炉模拟实验装置,通过该装置获得了发生冲顶时的红外热像。其次,通过对红外热图像的分析及不同区域温度曲线变化趋势,对缓慢冲顶、瞬时冲顶的形成过程,冲顶时不同区域温度曲线的变化规律进行了分析与总结。得到了通过红外热图像可以检测到造成冲顶事故的发生条件,可实现对发生冲顶时的预警功能,从而避免事故的发生。更多还原

【Abstract】 As the world’s energy crisis deepens and international oil price soars up. to search substitutes of oil becomes an urgent task for all countries. Oil shale resources are one of alternative energies of oil and natural gas. So use of oil shale to produce shale oil has drawn extensive attention from all countries again, including China where abundant reserves of oil shale are available. Control effects of those oil shale’s reactors have direct influence over outputs of shale oil while detecting and monitoring temperatures in reactors serve as one of key links in control systems of oil shale’s reactors, so it is of necessity to conduct in-depth studies on temperature monitoring of reactors in the process of producing oil from oil shale. This paper has monitored real-time charge level temperature changes in reactors and proposed a non-contact temperature measurement for oil shale’s reactors, by adopting infrared thermography, model-based image processing and temperature field modeling approach together with considerations of technological characteristics specific to oil shale’s reactors.It has studied destructive distillation of oil shale, analyzed and summarized various currently used temperature measurements for ring-shaped reactors, found out problems hindering temperature measurements for ring-shaped reactors. It also analyzes and studies effects of accurate temperature control in destructive distillation furnaces on oil shale’s oil productivity, mainly analyzes influence of temperature changes in destructive distillation furnaces on organic materials, shale oil and carbon residue and finds out optimal temperature control points to get highest-quality shale oil; this paper analyzes effects of temperature in destructive distillation furnaces on pyrolytic gas and obtains optimal temperature control points for maximum yields and highest-quality production of pyrolytic gas, providing the theoretical basis for optimal temperature control regarding different reaction stages in reactors.Reactors have complex internal environments and it is hard to extract charge level features in hot air developing status in rectors. To solve these problems, this paper analyzes charge level temperature field’s relations with charge level infrared heat pictures cross temperature measuring, and so on, coming up with a way to extract charge level temperature features in reactors.Modifications of C-V Model based on AOS format and temperature field partitioning algorithm coupled with background painting techniques (integrating background painting techniques, modified C-V Model and semi-implicit AOS) for infrared heat pictures which enables automatic segmentation of the temperature field in infrared heat pictures are proposed.Through considering margin-based and region-based active contour models, this paper puts forward an infrared thermography multiple-temperature-field region detection method based on a global optimum active contour model, enabling effective detection in infrared heat pictures’temperature fields and successful extraction of charge level temperature features. This resolves recognition difficulties caused by similar target and background colors plus unclear margins in temperature field regions under various shapes in infrared heat pictures (especially concave-shaped ones) and heat pictures.From the viewpoint of processes and mechanisms, it analyzes charge level temperature field’s relations with multi-source information including infrared heat pictures in reactors, cross temperature measuring and so on, as well as extraction techniques based on charge level temperature field’s isothermal lines and other features in infrared heat pictures. To solve the problem brought by difficulty to establish an accurate charge level temperature field model with use of single detecting information, this paper proposes a charge level temperature field modeling approach for reactors based on information fusion. Specifically, this approach fully uses reactors’cross temperature measuring information via two-point-based dynamic temperature calibrating of temperature fields. Simulation results show that this approach have favorable generalization ability and accuracy, confirming this approach’s feasibility and excellence.This paper sets up a charge level temperature field monitoring system based on information fusion by using the modeling approach proposed. This system uses a visualized interface to show the charge level temperature field model, presenting charge level temperature distributions in a more visual way and providing real-time reliable reference information to reactor operators.Regarding probability of lid-breaking in oil shale’s reactors, this paper conducts early warning studies in reactors’lid-breaking incidents and makes clear mechanisms underlying such incidences by using infrared thermal imaging techniques and considering temperature changes’ trends and laws in different regions of heat pictures. First, it establishes a simulation experiment facility for reactors with use of infrared thermography technologies and obtains infrared heat pictures through such facility in the event of lid breaking. Second, it analyzes and summarizes forming processes of slow and instant lid breaking plus change laws in different regions’ temperature curves when lid breaking occurs by studying infrared heat pictures and variation trends in different regions’temperature curves; by doing these, it obtains occurring conditions detectable in the event of lid breaking through infrared heat pictures, thus enabling early warning for lid breaking to avoid accidents.更多还原