【作者】 张大尉；

【导师】 王华强；

【作者基本信息】 合肥工业大学 ， 控制理论与控制工程， 2005， 硕士

【摘要】 高炉冶炼是在高温下发生的包含许多物理、化学变化传输过程的一个复杂过程。只有保持炉况稳定顺行,才能取得较好的技术经济指标。而炉况是经常波动的,因此对高炉异常炉况的预测和判断是当前高炉控制的主人问题。 目前的解决办法是建立高炉异常判断专家系统,但仍缺陷,如实时性较差,学习能力较差等。而神经网络由于其很强的学习能力和自适应性,被成功地应用于很多智能控制系统中。因此探索建立神经网络高炉异常炉况判断专家系统是当前的研究方向。本课题旨在建立神经网络高炉异常炉况判断专家系统,并解决传统专家系统在知识获取方面的“瓶颈”问题,使其具有在线学习能力。 本文研究了当前高炉控制检测和控制的现状,在对高炉异常炉况判断做了深入研究的基础上,采用SIEMENS公司的S7-300系列的PLC和工业控制计算机(IPC)构成了二级计算机监控系统。以Viscual C++6.0为开发工具编制了具有在线学习能力的神经网络高炉异常炉况判断专家系统的软件,此系统使用方便,速度快,命中率高,具有很强的实用性。更多还原

【Abstract】 Smelting in blast furnace is a complicated process under temperature including a lot of physical change , chemical change and transmission process . Except for keeping the smooth operation state on BF, better technique and economic norms is just obtained . But the state of BF is often undulate , so the main problem in current blast furnace.The current solution is to set up the expert system for predicting and judging the abnormal state of blast furnace , but is still has drawback.. For example , the character of real-time and the ability of learning are bad , and so on. But because neural network has very strong the ability of learning and adaptility , it has been used in a lot of control system successfully .It is the current research direction to set up the neural network expert system of judging the abnormal state of blast furnace . The object of this research is to set up the neural network expert system of judging the abnormal state in acquiring knowledge , so to make it have the ability of self-learning on-line.VisualC++6.0 with the strong programming ability are used to draw up the software of neural network expert with self-learning on-line ability for judging the abnormal state of blast furnace . According to the claim of the blast furnace to monitor system , we adopt two level monitor system with SIEMENS S7-300 series and industry control computer (IPC). The percentage of hits of this system is high and the speed is fast . Except these , this system is convenient in use . It cuts down the hardware resource of learning out-line and strengths the practicability.更多还原