首页 | 本学科首页   官方微博 | 高级检索  
     

基于模式识别的合成氨统计过程控制
引用本文:阎镜予,沈之宇,孙德敏. 基于模式识别的合成氨统计过程控制[J]. 化工自动化及仪表, 2006, 33(6): 31-34
作者姓名:阎镜予  沈之宇  孙德敏
作者单位:中国科学技术大学,自动化系,合肥,230027;中国科学技术大学,自动化系,合肥,230027;中国科学技术大学,自动化系,合肥,230027
摘    要:合成氨生产过程因影响因素多且关系复杂,容易引发各类异常.利用模式识别方法,依据操作类和目标类之间的对应关系,提出"越类量"概念,并给出计算方法.对"越类量"利用统计过程控制方法进行跟踪,在线检测过程异常,在一套年产8×104t的氨合成系统实际投运中取得了满意的效果.

关 键 词:统计过程控制  模式识别  越类量  合成氨
文章编号:1000-3932(2006)06-0031-04
收稿时间:2006-05-08
修稿时间:2006-05-08

Statistical Process Control for Ammonia Synthesis Based on Pattern Recognition
YAN Jing-yu,SHEN Zhi-yu,SUN De-min. Statistical Process Control for Ammonia Synthesis Based on Pattern Recognition[J]. Control and Instruments In Chemical Industry, 2006, 33(6): 31-34
Authors:YAN Jing-yu  SHEN Zhi-yu  SUN De-min
Abstract:The concept of "slop-over value" is proposed based on pattern recognition clustering method.A method to calculate slop-over value is presented according to the relationship between operation class and objective class.The exception condition is detected on line by tracking slop-over value with statistical process control method.The proposed method has been applied in the exception detection of an ammonia synthesizer(80,000 tons per year) with satisfactory results.
Keywords:statistical process control  pattern recognition  slop-over value  ammonia reactor
本文献已被 CNKI 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号