射频升降机电源故障诊断算法研究

Rf Elevator Power Fault Diagnosis Algorithms

在射频升降机运行环境中,大部分电源都是通过母线连接在配电网络中的,配电网络中的支路很多,对射频升降机电源线路的电流带来较大程度的干扰,导致故障信号转换的电流信号存在较大程度的波动。传统故障诊断方法,无法对上述信号波动进行准确的分类,精度不高。提出了一种改进的射频升降机电源故障诊断算法。应用LS-SVM方法对射频升降机电源信号进行聚类处理,利用一种惩罚因子对信号分类过错进行约束,运用融合规则对故障特征进行融合,完成射频升降机电源故障定位。实验结果表明,改进后的算法有效提高了故障诊断的准确性。

In the radio frequency elevator operation environment, most of the power is connected to the distribution network through the bus bar. There are many branch circuits in the distribution network, bringing forward larger degree of interference to the current of the power circuits of the RF elevator and resulting in larger degree of fluctua- tion of the current signal for fault signal translation. Traditional algorithms for fault diagnosis can not sort the signal fluctuations mentioned above accurately. An improved RF elevator power fault diagnosis algorithm was proposed in this paper. The LS - SVM method was used to perform the clustering processing for RF elevator power signal, a penalty factor was utilized to restrict the signal sorting error, and the fusion principle was used to fuse the fault characteristics, so as to realize the rf elevator power fault location. Experimental results show that the improved algorithm can effectively improve the accuracy for fault diagnosis.