摘要
把神经网络技术与规则推理技术相结合 ,用于方案设计决策支持系统的推理机构中。首先根据领域规则构造出神经网络 ,再根据样本对神经网络进行学习训练。该系统有自学习功能 ,其网络的拓扑结构在学习过程中可以动态变化 ,用户可根据实际需要进行知识的更改和完善 ,用户界面方便实用。
The paper combined the neural network technique into the rule-based reasoning machine. A new integrated inference method was presented and used to build a scheme design support system for mechanic products. The neural network is constructed according to the domain rules and trained by the samples. The networks topology structure of the system can change dynamically with the self-learning function. Users can modify the knowledge easily.
出处
《高技术通讯》
EI
CAS
CSCD
2003年第1期81-84,共4页
Chinese High Technology Letters
基金
国家自然科学基金 ( 6 97740 32
79970 0 30
6 0 10 40 0 4)资助项目。