瞿寿德. 智能自动化的模式识别方法[J]. 工程科学学报, 1998, 20(4): 385-389. DOI: 10.13374/j.issn1001-053x.1998.04.018
引用本文: 瞿寿德. 智能自动化的模式识别方法[J]. 工程科学学报, 1998, 20(4): 385-389. DOI: 10.13374/j.issn1001-053x.1998.04.018
Ju Shoude. Pattern Recognition Approach to Intelligent Automation for Complex Industrial Processes[J]. Chinese Journal of Engineering, 1998, 20(4): 385-389. DOI: 10.13374/j.issn1001-053x.1998.04.018
Citation: Ju Shoude. Pattern Recognition Approach to Intelligent Automation for Complex Industrial Processes[J]. Chinese Journal of Engineering, 1998, 20(4): 385-389. DOI: 10.13374/j.issn1001-053x.1998.04.018

智能自动化的模式识别方法

Pattern Recognition Approach to Intelligent Automation for Complex Industrial Processes

  • 摘要: 对于那些无法用数学方程式表述的复杂生产过程,提出一种基于模式识别原理的智能自动化新方法.该方法的基本点在于:从信息的角度出发,采用特征变量来描述过程工况,再用时间序列分析法建立预测模型,从而实现工艺最优化.还介绍了专门为此目的研制的红外线CCD热成像装置,并给出了建立预测模型的神经网络算法.

     

    Abstract: A new intelligent automation method on the basis of pattern recognition is proposed in situations that no mathematical equations are available to depict complex industrial processes.From information point of view the essence of this method is first to describe a process by use of feature vaeiables,and then establish prdiction model via time series analysis to provide useful information for process optimization.An infrared CCD thermovision system is introduced to fulfill the goal mentioned above and in order to establish the prediction model an algorithm based on neural network is provided as well.

     

/

返回文章
返回