【摘要】 用模块化模糊神经网络模型进行参数学习专家系统的研究是根据不同的炉况类型及其相关变量将整个参数学习模型分解为几个结构相似的子模糊神经网络模型。其中，子神经网络模型能够模拟诊断推理过程，动态地进行特征数据的权重、模糊等级的界限值和规则可信度的调整。采用该方法对现场采集的数据进行实验，结果表明该方法具有较好的学习效果。更多还原

【Abstract】 When the parameter learning expert system was researched by modular neuro fuzzy network model, the whole model was decomposed into a few sub models of neuro fuzzy network with similar structure according to different abnormal statuses. The sub models can simulate the inference process of diagnosis system, and tune the weights, the membership functions of the feature data and the reliabilities of the rules. The results of the experiments using the data collected from a blast furnace show that the proposed method is feasible.更多还原