针对矿用通风机故障具有不确定性和复杂性的问题,利用风机的振动参数进行推理,提取常见故障振动信号的特征频谱值来组建及训练神经网络,以此建立诊断系统进行故障类型的识别。诊断结果与实际故障相符,表明基于模糊神经网络故障诊断方法能够快速准确地得到风机故障的特征和状态,增加了风机故障诊断的可靠性和实用性。 Aiming at the problem of the mine ventilator fault with uncertainty and complexity, the fan vibration parameters are used to reason and extract the characteristic spectrum values ​​of the common fault vibration signals to form and train the neural network, so as to establish the diagnostic system to identify the fault types . The result of the diagnosis is in accordance with the actual fault, which shows that the fault diagnosis method based on fuzzy neural network can quickly and accurately get the characteristics and status of fan fault, and increase the reliability and practicability of fan fault diagnosis.