论文部分内容阅读
目的讨论基于模糊产生式规则的故障诊断专家系统与神经网络相结合的问题，把推理网络同神经网络联系起来，使它能转换成神经网络．方法在转换中，把模糊产生式规则前提的置信度归结到神经网络的输入信息学习样本，把规则的置信度归结到神经网络的输出信息学习样本，并给出实例的具体实现过程．结果解决了模糊规则专家系统向神经网络转换问题，实现了基于神经网络的模糊快速推理诊断和知识自动获取．结论经实例验证，该方法可靠有效．利用神经网络的并行处理和自学习能力，能避免传统模糊推理的冲突，低效率和知识获取的瓶颈问题．
Aim To discuss the combination of fault diagnosis expert system and neural network based on fuzzy production rules, and to associate inference network with neural network so that it can be transformed into neural network. In the process of conversion, the confidence of the rules of fuzzy production rule is attributed to the input information learning samples of neural network, and the confidence of the rules is reduced to the output information learning samples of neural network. The concrete realization process of the example is also given. Results The problem of the conversion from fuzzy rules expert system to neural network was solved, and the fuzzy reasoning and knowledge acquisition based on neural network was realized. Conclusion The results show that this method is reliable and effective. The use of neural network parallel processing and self-learning ability, to avoid the traditional fuzzy reasoning conflict, low efficiency and knowledge acquisition bottlenecks.