首页|基于深度学习的齿轮箱故障预测方法

基于深度学习的齿轮箱故障预测方法

扫码查看
机器人已广泛应用于汽车涂胶生产线,其突发故障会对生产节拍与成本造成很大的影响.目前,机器人本体的齿轮箱故障一般采用事后维修,因此迫切需要实施预防性维护措施.针对当前齿轮箱故障预测困难的问题,通过传感器采集齿轮箱的状态信息,建立故障预测深度学习模型,识别出可能导致故障的异常模式,从而实现故障的预测.首先建立基于生成对抗网络(GAN)的多变量时间序列信息异常检测框架,通过改进损失函数增强生成器的收敛性;然后引入基于时间扭曲编辑距离(TWED)的重构误差计算方法,精确计算时间序列信号的差异;其次采用基于局部异常概率(LoOP)的异常评价方法,对每个数据点进行异常评分,提高检测的准确率;最后以某白车身涂装单元对方法的有效性进行了应用验证.
Gearbox Fault Prediction Method Based on Deep Learning
Robots are extensively utilized in automotive gluing lines,where their sudden failures can signifi-cantly impact production pace and costs.Presently,gearbox failures in robot bodies are typically addressed post-occurrence,necessitating the urgent implementation of preventive maintenance measures.To tackle the challenge of difficult gearbox fault prediction,sensors are employed to collect the gearbox's state infor-mation,and a deep learning model for fault prediction is established.This model aims to identify abnormal patterns that could lead to faults,thereby enabling fault prediction.Initially,a multivariate time series a-nomaly detection framework based on a Generative Adversarial Network(GAN)is developed,which en-hances the generator's convergence through an improved loss function.Subsequently,a reconstruction er-ror calculation method based on Time-Warped Edit Distance(TWED)is introduced to accurately compute differences in time series signals.Furthermore,an anomaly evaluation method based on Localized Probabil-ity of Anomaly(LoOP)is adopted for scoring anomalies at each data point,thereby improving detection ac-curacy.Finally,the method's effectiveness is applied and validated in a body-in-white painting unit.

gearboxfault predictionmultivariate time seriesgenerative adversarial networksreconstruction error

史天一、时轮、何其昌

展开 >

上海交通大学机械与动力工程学院,上海 200240

齿轮箱 故障预测 多变量时间序列 生成对抗网络 重构误差

2024

传动技术
上海交通大学

传动技术

影响因子:0.197
ISSN:1006-8244
年,卷(期):2024.38(1)
  • 23