【摘要】 本文介绍了模糊信息扩散理论及其在故障诊断中的应用，运用模糊信息扩散理论对某液压系统的振动参数进行信息扩散，得到七组典型状态样本，训练神经网络，实例证明：通过典型样本的建立，克服了大样本的矛盾性及学习时间长的缺点，该网络对系统状态的诊断效果很好。更多还原

【Abstract】 This paper applies the method of fuzzy information processing to the construction of typical samples in large sample fault diagnosis. Typical diagnosis samples of a fluid power system are constructed by the theory of information diffusion. The samples help to get short training and reasonable diagnosis results in a diagnostic meural network.更多还原