【作者】 单修迎；

【导师】 贾春玉；

【作者基本信息】 燕山大学 ， 机械设计及理论， 2007， 硕士

【摘要】 板带材生产在国民经济中占有十分重要的地位，广泛应用于汽车、家用电器和宇航技术等方面。随着社会的不断进步，用户对板带材产品的质量提出了越来越高的要求。板厚和板形是板带材质量的两大主要质量指标，到目前为止，板厚问题已基本上得到了完善的解决，而对于板形问题，一直没有得到满意的解决，板形问题已成为日益迫切的急需解决的问题。倾辊和弯辊是常用的两种板形控制手段，是消除一次和二次板形的主要手段。在实际中，广泛使用传统的PID模型进行控制，本文在此基础上，根据从板形数据中提取出的一次和二次板形信息，利用神经网络动态调整PID控制器的参数，将基于神经网络的模糊PID模型应用到倾辊和弯辊控制中，以提高传统PID模型对板形的控制能力。分段冷却也是常用的板形控制手段，生产中主要用来消除高次板形，它是一个及其复杂的过程，很难建立一个精确的数学模型，本文针对这一问题，利用板形数据中的高次板形信息，在线地对分段冷却模型进行模糊辨识，根据辨识结果求逆，得到动态的模糊控制器进行控制，建立了分段冷却自适应模糊控制模型，以提高高次板形的控制质量。本文最后在所建立的板形智能控制模型的基础上，编写了动态仿真软件，形象地展示了板形控制效果。仿真结果表明所建立的基于神经网络的模糊PID倾辊弯辊控制模型和分段冷却自适应模糊控制模型可靠性高、适应性强，提高了板形综合控制的精度，推动了板形控制的发展，具有重要的实用价值。更多还原

【Abstract】 Strip production is very important in national product, which is widely used in automotive vehicle, household appliance and space navigation technology etc. With the development of our society, more and more high-quality products are needed by customers. Thickness precision and flatness are two main quality targets. Up to now, the problem of thickness precision has been solved on the whole, however, the problem of flatness has not been solved satisfactorily and it is more and more urgent to solve the problem.Tilting roll and bending roll are two common flatness control means and they are main means of eliminating linear and quadratic flatness. In practice, conventional PID mode is used widely, based on which fuzzy PID based on neural network is used in tilting roll and bending roll control to raise flatness control ability of conventional PID mode using neural network to adjust the parameters of PID controller in the paper.Sub-sectional cooling is also common flatness control means, which is mainly used to eliminate higher degree flatness. It is a most complex process and it is hard to build a precise mathematical model, so self-adaptive fuzzy control effects for sub-sectional cooling model is built by identifying sub-sectional cooling model by higher degree flatness information of flatness data and gaining dynamic fuzzy controller to raise higher degree flatness control quality in the paper.Last in the paper, a dynamic simulating software is compiled base on the built flatness intellect control model, by which flatness control effects. Simulating results indicate that the built fuzzy PID based on neural network tilting roll and bending roll model and self-adaptive fuzzy control effects for sub-sectional cooling model have high reliability and strong adaptability, raising flatness integrated control precision, promoting the development of flatness control and having important practical value.更多还原