학위논문(석사)--서울대학교 대학원 :공과대학 기계항공공학부,2019. 8. 윤병동. ; As smart factory is becoming popular, industrial robots are highly demanding in many manufacturing fields for factory automation. Unpredictable faults in the industrial robot could bring about interruptions in the whole manufacturing process. Therefore, many methods have been developed for fault detection of the industrial robots. Because gearboxes are the main parts in the power transmission system of industrial robots, fault detection of the gearboxes has been widely investigated. Especially, vibration analysis is a well-established technique for fault detection of the industrial robot gearbox. However, the vibration signals from the gearboxes are mixed convolutively and linearly at each axes, which makes it difficult to locate a damaged gearbox, and reduce fault detection performance. Thus, this paper develops a vibration signal separation technique for fault detection of industrial robot gearboxes under multi-axis interference. The developed method includes two steps, frequency domain independent component analysis (ICA-FD) and time domain independent component analysis (ICA-TD). ICA-FD is aimed at separating convolutive mixture of signals, and ICA-TD is aimed at eliminating the residual mixed components. The experiment is performed to demonstrate the effectiveness of the proposed method. The results show that the proposed method could successfully separate the mixed signals by obtaining vibration signals from each gearbox, and enhance fault detection performance for the industrial robot gearboxes. ; Chapter 1. Introduction 1 1.1 Background and Motivation . 1 1.2 Scope of Research 1 1.3 Structure of the Thesis . 5 Chapter 2. Structure of Industrial Robot . 6 2.1 Structure of Experimental Robot 6 2.2 Problem in Industrial Robot Fault Detection . 8 Chapter 3. Methodology 10 3.1. Time Domain Independent Component Analysis (ICA-TD) . 10 3.2. Frequency Domain Independent Component Analysis (ICA-FD) 12 3.2.1 Separation 12 3.2.2 Permutation . 14 3.2.3 Scaling . 17 3.3. Multi-stage Independent Component Analysis (MSICA) . 17 Chapter 4. Experiment Evaluation . 19 4.1 Experiment with MSICA 19 4.1.1 Experiment Process . 19 4.1.2 Result Analysis 28 4.2 Comparison Experiment Using Basic ICA Method . 33 4.3 Comparison Experiment Using ICA-FD Method . 38 Chapter 5. Discussion and Conclusion . 45 5.1 Conclusions and Contributions 45 5.2 Future Work 46 Bibliography . 47 ; Master
Minimization of Multi-Axis Interference for Fault Detection of Industrial Robots Based on Blind Source Separation ; 산업용 로봇 고장 진단을 위한 암묵신호 분리 기반 다축 간섭 최소화 기법
2019-01-01
000000000040▲000000000041▲000000157774▲
Theses
Electronic Resource
English
DDC: | 629 |
Bearing Fault Diagnosis of the Gearbox Using Blind Source Separation
SAE Technical Papers | 2020
|British Library Conference Proceedings | 2012
|Blind source separation of images based upon fractional autocorrelation
British Library Online Contents | 2012
|