基于DS-DBN多测点的工业机器人RV减速器故障诊断分析

Fault Diagnosis of RV Reducer of Industrial Robot Based on DS-DBN Multiple Measurement Points

为了提高减速器故障识别能力,融合DS证据理论和深度信念网络(DBN)设计了一种基于DS-DBN多测点的故障诊断方法,并开展实验测试验证。研究结果表明:经过30代训练后达到收敛状态,训练四个测点的DS-DBN故障诊断模型分类误差均小于5%,该方法能够快速获得全局最优参数。相比较单点测试,通过DS证据理论完成测点故障诊断融合处理后,可以将故障诊断准确率提升至接近100%的程度。该研究有助于解决工业机器人RV减速器隐藏的问题,延长使用寿命、减少经济成本。

RV gearboxes are widely used in industrial robots,which directly affect the operating efficiency and cost budget of industrial robots.In order to improve the fault identification ability of the reducer,a DS-DBN multi-measurement point based fault diagnosis method is designed by integrating DS evidence theory and deep belief network(DBN),and experimental tests are carried out to verify it.The results show that the convergence state is reached after 30 generations of training,and the classification errors of the DS-DBN fault diagnosis model for training four measurement points are less than 5%,indicating that the global optimal parameters can be obtained quickly by using the method in this paper.Compared with the single-point test,the fault diagnosis accuracy can be improved to nearly 100%after completing the measurement point fault diagnosis fusion processing by DS evidence theory.This research helps to reject the hidden problems of RV reducer of industrial robots and improve the service life and economic cost.