【作者】 游廉明；

【导师】 阮学斌；

【作者基本信息】 福州大学 ， 系统工程， 2006， 硕士

【摘要】 本文介绍了智能控制在料场配料控制系统中的具体应用。针对配料过程中不同下料圆盘对象的控制要求,分别采用免疫模糊控制和神经网络预测控制进行理论研究,并在实际应用中取得了良好的控制效果。 论文的主要研究工作和成果包括以下几个方面: 1、简要介绍了智能控制技术的产生、发展、现状及其发展趋势以及智能控制技术中的模糊控制技术和神经网络控制技术的发展和现状。 2、本课题是福建三明闽光钢铁集团有限公司烧结分厂原料场2004-2005年计算机监控系统技术改造项目,在此将智能控制技术应用于智能化圆盘称重配料系统的控制中。 3、介绍了免疫模糊PID控制的原理和设计方案。 4、介绍了神经网络控制的原理和设计,重点介绍了径向基神经网络的特点和设计方法,以及径向基神经网络在模型预测控制中的应用。 5、介绍了三钢烧结称重配料计算机监控系统的系统软硬件结构和实现功能。 6、针对不同下料圆盘的控制要求,采用模糊PID控制方案和基于径向基神经网络预测控制方案对控制系统进行设计和仿真研究,表明了系统的可行性和有效性,同时也在实际应用取得了良好的效果。更多还原

【Abstract】 Application of intellectual control technology in the raw material distribution system is dealt with. According to the control request of the different chutes of whole system, immune fuzzy control and neural network control are respectively adopted to be studied theoretically. Finally, the good control result in practical application has been reached.The main research work and contributions of this dissertation list as follows:1. A survey of the production 、 development、 current situation and future trend of intellectual control technology is summarized in brief, especially for fuzzy control technology and neural network control technology .2. The study in this dissertation is a part of the year 2004 to 2005 technology reformation Project of Fujian Sanming MingGuang Steel Group Fire Filiale. Here, intellectual control technology is applied to control the raw material distribution system.3. An introduction of the principle and design of immune fuzzy PID control is given.4. An introduction of the principle and design of neural network control is given. Especially, the character and learning arithmetic of Radial Basis Function neural network and the application of Radial Basis Function neural network in the model predictive control are introduced emphatically.5. An introduction of system structure and function of computer monitoring system for raw material distribution system.6. In order to meet the control requirement of the different chutes, immune fuzzy PID control and Model predictive control based on Radial Basis Function neural network strategy is adopted to carry on designing and simulation research. And consequently, systematic feasibility and validity is confirmed, and the actual application also achieves good effect.更多还原