摘要
本文试用模式识别方法研究甲醛生产过程,并通过按PCA方法选择稳定子集,建立了主分量y_1和y_2的平面而获得了该系统的控制模型,求得了在满足产品质量和产量约束条件下使原料单耗最低的优化操作条件.由实际运行表明,本文提供的最优条件可显著地降低原料单耗,提高产品质量,延长催化剂寿命.
Pattern recognition is applied to the research of formaldehyde production process. By using PCA method, a stable subset is selected and a plane with the principal components y_1 and y_2 is established, thus a control model of the process and an optimum operation condition of making the lowest raw material consumptions and satisfying the constrained conditions of the product quality and output are obtained. Practical operations show that the optimum conditions presented in this paper can greatly reduce the raw material consumptions, increase product quality and prolong catalyst life.
出处
《自动化学报》
EI
CSCD
北大核心
1989年第2期161-164,共4页
Acta Automatica Sinica
关键词
模式识别
甲醛
最优化
计算机控制
Formaldehyde
computer control
optimization
pattern recognition
