摘要
DMOS软件综合运用了模式识别、支持向量机、人工神经网络、遗传算法、线性和非线性回归等多种数据挖掘技术,能有效解决复杂工业过程系统优化中普遍存在的多因子、高噪声、非线性、非高斯分布和非均匀分布的难题。将DMOS工业优化软件成功地应用于柴油加氢精制装置及丙烯腈反应装置的生产优化。根据装置DCS系统采集的生产数据,研究了装置优化操作的主要工艺参数,采用模式识别方法建立了装置生产优化操作的定性模型,并最终建立了优化目标的数学模型。
Many kinds of data mining technology that include pattern recognition, support vector machine, artificial neural network ,genetic algorithm ,linear or nonlinear regression method are applied in DMOS( Data Mining Optimization System). The DMOS can effectively solve the problems which have multiple-factor, high noise, nonlinear, nonGaussian distribution and non-uniform distribution in complicated industry process optimization. The DMOS software is successfully applied to industrial process optimization of the diesel oil hydrofining unit and the acrylonitrile reactor unit. According to the process data gathered by DCS, the unit major optimization operation process parameters are researched. Using pattern recognition method, the qualitative model of the unit production optimization operation is built. The mathematical model of optimization objective is established.
出处
《化工自动化及仪表》
EI
CAS
2005年第4期36-39,共4页
Control and Instruments in Chemical Industry
基金
国家自然科学基金资助项目(20373040)
宁波市重点博士(青年)基金资助项目(2003A61005)
关键词
DMOS
数据挖掘
模式识别
过程控制与优化
加氢精制
丙烯腈
DMOS
data mining
pattern recognition
process control and optimization
hydrofining
crylonitrile.