论文部分内容阅读
论述了建立规则型模糊神经网络的理论和方法，针对大型旋转机械提出了一种采用多层规则库结构及智能推理机的故障诊断技术。该技术以 Ｒｕｌｅ 型模糊联想记忆器作为诊断系统的分类和综合算法，把基于知识的符号处理方法与模糊神经网络有机地结合在一起。讨论了模糊神经网络输入和输出模糊化的问题。为电厂汽轮发电机组故障诊断专家系统提供了新的思路。
The theory and method of establishing regular fuzzy neural network are discussed. Aiming at large rotating machinery, a fault diagnosis technology based on multi-layer rule base and intelligent inference engine is proposed. This technology uses Rule-type fuzzy associative memory as a classification and synthesis algorithm for diagnostic systems, and organically combines knowledge-based symbol processing with fuzzy neural networks. Discusses the fuzzy neural network input and output fuzzification problems. It provides a new idea for the fault diagnosis expert system of power plant steam turbine.