基于深度学习的机器人电气故障检测与诊断研究
Research on Robot Electrical Fault Detection and Diagnosis Based on Deep Learning
农钧麟1
作者信息
- 1. 广西职业技术学院,广西南宁 530226
- 折叠
摘要
机器人以其高效、高强度的作业方式,被广泛地应用于制造业及其相关领域.但当机器人出现故障时,通常会导致生产线停滞,浪费大量的人力、物力,甚至危及工作人员的生命安全.传统的故障诊断方法耗时长,诊断效率低,且故障辨识精度不高.文章基于深度学习相关理论,通过分析机器人机械臂各个关节及执行器的振动特性,建立了一种适用于工业机器人的故障诊断模型,并以ABBirb120机器人为例进行电气故障检测与诊断的准确率分析.结果表明,在迭代次数达到900以上时,其故障识别准确率趋于99.4%.
Abstract
Robots are widely used in manufacturing and related fields due to their efficient and high-intensity operation methods.However,when robots malfunction,it often leads to production line stagnation,wasting a lot of manpower and material resources,and even endangering the safety of workers.Traditional fault diagnosis methods are time-consuming,inefficient,and have low accuracy in fault identification.Based on deep learning theory,this article establishes a fault diagnosis model suitable for industrial robots by analyzing the vibration characteristics of various joints and actuators of robotic arms.The accuracy of electrical fault detection and diagnosis is analyzed using the ABBirb120 robot as an example.The results show that when the number of iterations reaches 900 or more,the accuracy of fault identification tends to 99.4%.
关键词
深度学习/机器人/电气故障/检测Key words
deep learning/robots/electrical failure/detection引用本文复制引用
出版年
2024
NETL
NSTL
万方数据