摘要
文中分析了聚合物微观质量控制系统对模式识另模块提出的任务与要求，并在此基础上确定了人工神经元网络的结构．作者阐述了利用采集的系统数据对神经元网络进行训练和验证的过程，及首次应用人工神经元网络实现聚合物微观质量控制系统中模式识别的功能．
The task and requirements of the pattern recognition module pertaining to the microscopic quality control system for polymers(MQCSP) have been analysed and,based upon which,a structure of artificial neural networks(ANN) has been selected.Following an account of the procedure of training and substantiating the ANN with sampled data of the system,the author's experience in the maiden application of ANN to the materialisation of pattern recognition for the MQCSP has neen depicted.
关键词
人工神经网络
质量控制
高聚物
molecular weight distribution
pattern recognition
artificial neural networks
microscopic quality control
