华南理工大学学报(自然科学版) ›› 2009, Vol. 37 ›› Issue (2): 14-19.

• 机械工程 • 上一篇    下一篇

基于知识的产品制造过程能耗的计算与预测

宫运启 吕民 王刚 付宜利   

  1. 哈尔滨工业大学 机电工程学院, 黑龙江 哈尔滨 150001
  • 收稿日期:2008-05-15 修回日期:2008-09-01 出版日期:2009-02-25 发布日期:2009-02-25
  • 通信作者: 宫运启(1971-),男,博士生,主要从事先进制造技术、知识管理等研究. E-mail:gongyunqil@sohu.com
  • 作者简介:宫运启(1971-),男,博士生,主要从事先进制造技术、知识管理等研究.
  • 基金资助:

    黑龙江省自然科学基金资助项目(E200634)

Knowledge-Based Calculation and Prediction of Energy Consumption in Product Manufacturing Process

Gong Yun-qi  Lu Min  Wang Gang  Fu Yi-li   

  1. School of Mechatronics Engineering, Harbin Institute of Technology, Harbin 150001, Heilongjiang, China
  • Received:2008-05-15 Revised:2008-09-01 Online:2009-02-25 Published:2009-02-25
  • Contact: 宫运启(1971-),男,博士生,主要从事先进制造技术、知识管理等研究. E-mail:gongyunqil@sohu.com
  • About author:宫运启(1971-),男,博士生,主要从事先进制造技术、知识管理等研究.
  • Supported by:

    黑龙江省自然科学基金资助项目(E200634)

摘要: 为了估算新产品制造过程中的能耗,提出了基于知识的能耗预测方法.首先,应用本体技术使能消的计算和管理更清晰和明确,建立了能耗知识的语义模型并指出其语义表示方法.其次,提出了针对能耗预测特点的工艺实例检索过程及相似实例结果的重用方法.最后,应用神经网络技术建立了针对加工参数的能耗预测模型,并以叶片加工为例进行了应用验证.从而形成了从全局视角到工艺过程再到工序的能耗预测层次框架,并对原型系统的开发及应用进行了说明.

关键词: 能耗预测, 离散制造业, 本体, 实例推理, 神经网络

Abstract:

To estimate the energy consumption in the manufacturing process of a new product, a knowledge-based prediction method is proposed in this paper. In this method, first, the ontology technology is used to explicate the calculation and management of energy consumption, and a semantic model of energy consumption knowledge and its semantic representation are proposed. Then, a case retrieval process and a method of ease result reuse are presen- ted in view of the characteristics of energy consumption. Moreover, an energy consumption prediction model corre- sponding to manufacturing parameters is established by means of neural network. The effectiveness of the model is finally verified by the machining process of a blade. Thus, a hierarchical framework of energy consumption predic- tion from the overall view to the technical process and further to the machining precedure is formed. The develop- ment and application of the corresponding prototype system are also illustrated in the paper.

Key words: energy consumption prediction, discrete manufacturing industry, ontology, ease-based reasoning, neural network