Fault Diagnosis of Quadruped Robot Joint Modules Based on TDIC and Random Forest International Conference on Sensing, Measurement & Data Analytics in the era of Artificial Intelligence (ICSMD2024)

Presentation Details

Fault Diagnosis of Quadruped Robot Joint Modules Based on TDIC and Random Forest

ID：62 View Protection：ATTENDEE Updated Time：2024-10-23 10:43:36 Hits：667 Oral Presentation

Start Time：2024-11-01 16:20（Asia/Shanghai）

Duration：20min

Session：P3 Parallel Session 3 » P3-1Parallel Session 3（November 1 PM）

No files

Abstract

摘要—随着机器人技术的发展，四足机器人在救援任务和医疗保健等各个领域发挥着重要作用。然而，由于工作条件复杂，四足机器人关节模块中的组件经常受到交变和冲击载荷的影响。因此，研究关节模块中行星减速机的信号处理和故障诊断技术具有重要的工程意义。考虑到行星减速机故障信号提取难度高、难度大、非平凡性，提出了一种基于时变本征相关（TDIC）的故障特征提取方法，并结合随机森林分类器进行故障分类。该方法基于信号频率分析，自适应地提取不同物理场信号之间相关性高的片段作为信号特征。采用随机森林模型对行星齿轮箱中的故障进行分类，并与 CNN、KNN 和 XGBoost 模型进行比较。实验结果表明，所提方法具有优异的故障分类性能。

Keywords

planetary gearbox fault diagnosis,Time-Dependent Intrinsic Correlation,feature extraction,Quadruped Robot,Random Forest

Speaker

ChenYuFan

Master Degree Candid Shanghai University

Submission Author

ChenYuFan Shanghai University

FanBeibei Shanghai Key Laboratory of Intelligent Manufacturing and Robotics

XiongXin Shanghai Key Laboratory of Intelligent Manufacturing and Robotics

ShiMaoping Shanghai University

Submit Comment

All Comments

Important Date

Conference Date

Oct 31

2024

to

Nov 03

2024
Sep 30 2024

Draft paper submission deadline
Nov 12 2024

Registration deadline

Contact Information

Mrs. icsmd
ic******@163.com

International Conference on Sensing, Measurement & Data Analytics in the era of Artificial Intelligence ((ICSMD2024))