摘要
将神经网络与人工智能技术相结合，提出了１种新型的知识表示的形式化描述和知识单元。知识通过模式分类导入，并用ＩＭＳＢＰ和（Ｂ－Ｂ）ＢＰ调节２种不同类型的权。在此基础上提出了神经网络的双向求解策略。以加热炉控制验证了其可行性。
The method combining the neural network with the traditional expert system is studied.A new description language for neural network knowledge representation and the concept of knowledge unit is proposed. The knowledge guides in neural network by (B-B) BP and IMSBP. And a Solving Strategy for the neural network positive and negative problem is presented. A example of the heating furnace controlled object by using the description language is given.
出处
《北京科技大学学报》
EI
CAS
CSCD
北大核心
1996年第2期158-162,共5页
Journal of University of Science and Technology Beijing
基金
国家自然科学基金
