针对当前矿井通风机机械故障诊断所面临的问题,提出了一种粗糙集-遗传神经网络分类器模型和它的构造方法,模型先利用粗糙集理论约简样本决策表属性,然后再利用遗传神经网络进行网络训练。通过与基本BP网络模型的对比,验证了该方法用于故障诊断的有效性。 Aiming at the problems of mechanical fault diagnosis of mine ventilator, a rough set-genetic neural network classifier model and its construction method are proposed. The model first uses the rough set theory to reduce the attributes of the sample decision table and then uses the genetic neural Network training network. Compared with the basic BP network model, the effectiveness of this method for fault diagnosis is verified.