根据多传感器数据融合理论,研究了煤矿大型机电设备故障类型,提出了用多传感器数据融合技术建立2级机电设备故障诊断的方法。在特征级上采用遗传算法优化BP神经网络(GA-BP),对大量多传感器监测数据进行降维并建立基本信任分配函数;在决策级采用D-S证据理论算法,克服故障诊断过程的不确定性,以提高故障诊断的准确性。 According to the multi-sensor data fusion theory, the fault type of large-scale mechanical and electrical equipment in coal mine is researched. A fault diagnosis method based on multi-sensor data fusion technology is proposed. Genetic algorithm is used to optimize the BP neural network (GA-BP) on the feature level to reduce the dimension of a large number of multisensor monitoring data and establish the basic trust distribution function. At the decision level, the DS evidence theory algorithm is used to overcome the uncertainty of the fault diagnosis process , To improve the accuracy of fault diagnosis.