论文部分内容阅读
为提高煤矿井下供电系统的可靠性,在不同电压、电流、功率因数、环境相对湿度条件下,开展了因机械振动引发的串联型故障电弧模拟实验。分析了不同实验参数对故障电弧的影响;提取串联型故障电弧相邻五周期电流信号中的过零点数、归一化后的方差、协方差构成特征向量;建立了基于随机森林分类算法的串联型故障电弧诊断模型,以正常运行及故障电弧电流信号的特征向量构成训练样本和测试样本作为随机森林模型的输入,对样本进行分类,进而诊断是否发生串联型故障电弧。结果表明,该方法能够有效地实现矿用电连接器串联型故障电弧的诊断。
In order to improve the reliability of underground power supply system, a series of fault arc simulations triggered by mechanical vibration were carried out under different voltage, current, power factor and environmental relative humidity. The influence of different experimental parameters on the fault arc was analyzed. The zero crossing points in the adjacent five-cycle current signals of the fault arc were extracted, and the normalized variance and covariance were used to construct the eigenvectors. A series of stochastic forest classification algorithms Type fault arc diagnosis model, the training samples and test samples are constructed as the input of the random forest model by the eigenvectors of the normal running and fault arc current signals, and the samples are classified to diagnose whether a series fault arc occurs. The results show that the proposed method can effectively diagnose tandem fault arc of mine electrical connectors.