介绍了一种加权支持向量机的原理和方法,建立了矿井风机故障检测模型。通过提取风机振动信号中的故障特征参数,引入类权重因子和样本权重因子,解决训练样本类别数量和重要程度不平衡而导致的错分问题,实现了对矿井风机故障的在线检测。 The principle and method of a weighted support vector machine are introduced, and the fault detection model of mine fan is established. By extracting fault characteristic parameters of fan vibration signals and introducing class weight factors and sample weight factors, the problem of misclassification caused by the imbalance of the number of training samples and the degree of importance is solved, and the on-line detection of mine fan failure is realized.