计算机应用 ›› 2005, Vol. 25 ›› Issue (07): 1645-1646.DOI: 10.3724/SP.J.1087.2005.01645

• 人工智能 • 上一篇    下一篇

半导体生产线的RBF神经网络预测建模

王令群1,2,潘石柱1,郑应平1,2   

  1. 1.同济大学 控制科学与工程系,上海 200092; 2.机械制造系统工程国家重点实验室,陕西 西安 710049
  • 收稿日期:2004-12-13 修回日期:2005-04-28 发布日期:2005-07-01 出版日期:2005-07-01
  • 作者简介:王令群(1979-),女,山东邹城人,博士研究生,主要研究方向:半导体生产线调度、智能控制;潘石柱(1976-),男,河南信阳人,博士研究生,主要研究方向:图像处理、模式识别;郑应平(1941-),男,研究员,博士生导师,主要研究方向:复杂系统、智能控制
  • 基金资助:

    国家973项目(20002CB31220203);国家自然科学基金资助项目(60374005,60343002)

Predictive model of semiconductor manufacturing line based on RBF neural network

WANG Ling-qun1,2,PAN Shi-zhu1,ZHENG Ying-ping1,2   

  1. 1. Department of Control Science and Engineering, Tongji University; 2. State Key Laboratory for Manufacturing Systems Engineering
  • Received:2004-12-13 Revised:2005-04-28 Online:2005-07-01 Published:2005-07-01

摘要:

预测模型是预测控制的基础。半导体生产过程的复杂性和随机性,使之难以建立确定性模型。提出一种新方法,利用径向基(RBF)神经网络对该过程建立预测模型。使用simul8软件对之在各种投料策略和调度策略下进行仿真并定时采样,将采样得到的数据对模型进行训练测试,结果表明该模型的预测输出与实测样本基本吻合,网络模型具有很好的泛化能力。训练后的网络可以对半导体生产线进行快速准确的预测,为预测控制和实时调度打下基础。

关键词: RBF神经网络, 半导体生产线, 预测控制

Abstract:

Semiconductor manufacturing process's complexity and randomness make it difficult to build determinate prediction model. A new method was presented which used RBF neural network to model this process. Manufacturing lines with various release control and scheduling policies were simulated by software simul8, and the sampling data got from the simulation model were used in the training and test of the prediction model. Results demonstrate that the model's output and the real samples output are basically identical and the model has great generalization ability. So the well-trained network can be used to forecast the state of the process rapidly and accurately, which lays foundation to prediction control and real time scheduling.

Key words: RBF neural network, semiconductor manufacturing line;, prediction control

中图分类号: