摘要
本文基于［１］所述知识流推理的思想，根据模糊控制的近似推理模型建立了一种动态模糊推理神经网络结构。该动态网络结构的推理过程是收敛的，文中给出了关于收敛性的充分性定理及理论分析。通过对该网络分析表明，该网络很适合于模糊知识的快速并行处理，并且对复杂近似推理的逻辑结构能直接通过该网络表达。
According to the main idea of knowledge flow inference presented in literature[1],a Dynamic Fuzzy Inference Neural Network(DFINN ) Model is established in this pa-per,which is based on the approximate inference model of fuzzy control.The inferenceprocess of the dynamic network is convergent.By the analysis and the application of theneural network,it is shown that the presented network is suitable for fast parallel pro-cessing of fuzzy knowledge,and the direct realization of complex fuzzy approximate in-ference logics.
出处
《模糊系统与数学》
CSCD
1994年第1期62-73,共12页
Fuzzy Systems and Mathematics
基金
四川省青年科技基金
