摘要
针对传统人工智能专家系统的缺点,引入了以并行计算为主的神经网络技术,并成功地运用它来建造工艺设计专家系统。神经网络中的BP(BackPropagation)模型用作规则记忆器,每一模型中,知识库和推理机是合二为一的,许多这类模型构成系统的知识库。
This paper introduces neural network technique based on parallel calculation and uses it to build an expert system of process design successfully. The system can overcome the shortcomings of traditional artificial intelligence expert system. BP (Back Propagation) model of neural network is regarded as the rule memory. Knowledge base and inference strategy of BP model are combined into one. Many BP models form the knowledge base of system. The forward inference of data driving realizes the final decision of process route.
出处
《农业机械学报》
EI
CAS
CSCD
北大核心
1996年第3期120-123,共4页
Transactions of the Chinese Society for Agricultural Machinery
关键词
工艺设计
专家系统
神经网络
Process design, Expert system, Neural network