基于模糊神经网络的泡生法蓝宝石长晶炉的智能控制系统设计

发布时间：2018-02-20 00:45

本文关键词： 泡生法蓝宝石晶体模糊神经网络智能控制　出处：《浙江理工大学》2017年硕士论文　论文类型：学位论文

【摘要】：人造蓝宝石晶体是以高纯度的Al2O3粉末为原料,加热熔化并通过籽晶降温提拉使其再结晶的过程。泡生法是人造蓝宝石晶体的一种方法,也是国内生产蓝宝石晶体的主流方法。目前国内泡生法蓝宝石晶体生长设备大多都是人工控制,需花费大量人力资源,并且人工水平参差不齐,不能保证每一炉晶体的质量。将泡生法蓝宝石生长设备与智能控制算法相结合,在降低生产成本的同时还可稳定晶体生长质量,是目前行业发展的重点所在。本文以模糊神经网络算法为基础,研究设计一种模糊神经网络控制器并将其与泡生法蓝宝石长晶炉相结合,从而实现自动控制泡生法蓝宝石长晶。模糊神经网络继承了模糊算法与神经网络各自的优点,可实现对任意非线性系统的逼近。本文主要研究内容和研究结果如下:1、通过对泡生法蓝宝石长晶炉的长晶过程以及设备结构的研究,分析了各个变量对长晶过程的影响,选取合适的控制变量,并根据自身实际操作经验与相关专家经验,对泡生法蓝宝石长晶过程进行归纳,得出相应模糊规则。以此为基础设计了模糊神经网络控制器。2、对所设计的模糊神经网络控制器进行仿真,以实际采集到的生长数据对控制器进行训练,并用训练完成的控制器进行预测,对控制器的效果进行验证,分析该模糊神经网络控制器的可行性。3、以所设计的模糊神经网络控制器为基础,从硬件选型方面着手,逐步搭建设备,并编写上位机控制软件,将控制器应用到实际生产过程中,对所设计的控制器控制效果进行实际验证。本文设计了以模糊神经网络算法为基础的控制器,并从理论仿真及实际应用两个角度对该控制器的控制效果进行了验证,在人造蓝宝石自动化领域有一定的价值。
[Abstract]:The synthetic sapphire crystal is a process of heating and melting with high purity Al2O3 powder and recrystallization by lowering the temperature of seeded sapphire crystal, which is a kind of method of synthetic sapphire crystal. It is also the mainstream method of producing sapphire crystals in China. At present, most of the domestic equipment for the growth of sapphire crystals are controlled manually, requiring a large amount of human resources and varying levels of manpower. The quality of every crystal can not be guaranteed. By combining the Bubble Sapphire growth equipment with the intelligent control algorithm, the quality of crystal growth can be stabilized while reducing the production cost. This paper studies and designs a fuzzy neural network controller based on fuzzy neural network algorithm and combines it with sapphire long crystal furnace. The fuzzy neural network inherits the advantages of fuzzy algorithm and neural network. The main contents and results of this paper are as follows: 1. By studying the long crystal process and the equipment structure of the sapphire long crystal furnace, the influence of various variables on the long crystal process is analyzed. Selecting appropriate control variables, and according to its own practical operation experience and relevant expert experience, the long crystal process of Bubble Sapphire method is summarized. The corresponding fuzzy rules are obtained. Based on this, the fuzzy neural network controller .2is designed, and the designed fuzzy neural network controller is simulated, and the controller is trained by the actual growth data. The effect of the fuzzy neural network controller is verified, and the feasibility of the fuzzy neural network controller is analyzed. Based on the designed fuzzy neural network controller, the hardware selection of the controller is introduced. The equipment is built step by step, and the control software of the upper computer is written. The controller is applied to the actual production process, and the control effect of the designed controller is verified. The controller based on the fuzzy neural network algorithm is designed in this paper. The control effect of the controller is verified from two aspects of theoretical simulation and practical application, and it has certain value in the field of artificial sapphire automation.
【学位授予单位】：浙江理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TQ164.2;TP183

【参考文献】