引用本文:梁炎明,刘丁,赵跃.基于自适应噪声抵消的CZ单晶炉炉膛温度信号处理[J].控制理论与应用,2011,28(1):94~100.[点击复制]
LIANG Yan-ming,LIU Ding,ZHAO Yue.Temperature signal processing based on adaptive noise cancellation for CZ crystal growing furnace[J].Control Theory and Technology,2011,28(1):94~100.[点击复制]
基于自适应噪声抵消的CZ单晶炉炉膛温度信号处理
Temperature signal processing based on adaptive noise cancellation for CZ crystal growing furnace
摘要点击 2486  全文点击 2133  投稿时间:2009-11-18  修订日期:2010-04-16
查看全文  查看/发表评论  下载PDF阅读器
DOI编号  10.7641/j.issn.1000-8152.2011.1.CCTA091466
  2011,28(1):94-100
中文关键词  CZ单晶炉  炉膛温度信号  低频干扰  滤波  傅立叶级数  自适应噪声抵消  粒子群优化
英文关键词  CZ crystal growing furnace  furnace temperature signal  low-frequency interference  filtering  Fourier series  adaptive noise cancellation  PSO
基金项目  国家科技重大专项资助项目(2009ZX02011001).
作者单位E-mail
梁炎明* 西安理工大学 自动化与信息工程学院 liangym@xaut.edu.cn 
刘丁 西安理工大学 自动化与信息工程学院  
赵跃 西安理工大学 自动化与信息工程学院  
中文摘要
      为滤除CZ单晶炉炉膛温度信号在单晶生长过程中存在的低频干扰, 提出一种基于噪声抵消技术的滤波方法. 首先利用傅立叶级数构造出低频干扰的逼近信号, 然后根据炉温信号的缓变特征建立能够获取低频干扰逼近信号的抵消器误差函数, 最后利用一种改进的粒子群优化算法优化误差函数获得低频干扰的逼近信号, 并用该逼近信号抵消低频干扰. 实验结果表明, 所提出的自适应噪声抵消滤波算法能够有效滤除CZ单晶炉炉膛温度信号中的低频干扰, 并优于常用的滤波方法.
英文摘要
      To filter the low-frequency interferences from the temperature signal in the single crystal growth process within a CZ crystal growing furnace, we propose a filtering method based on adaptive noise cancellation. Firstly, the approximation signals of the low-frequency interferences are modeled by using Fourier series. Secondly, based on the slow variation characteristics of the furnace temperature signal, the error function of cancellation is developed. Finally, the error function is minimized with an improved PSO (particle swarm optimization) algorithm to obtain the approximate signals of the low frequency interferences. These approximate signals are employed to offset the low-frequency interferences. Experiments show that this filtering method based on adaptive noise cancellation can efficiently attenuate the low-frequency interferences in the temperature signal of CZ crystal growing furnace, and is better than the most commonly-used methods in numerical precision.