Industrial Robot Rotate Vector Reducer Fault Detection Based on Hidden Markov Models

Zhang, Yinlong; An, Haipeng; Ding, Xiaoxi; Liang, Wei; Yuan, Mingze; Ji, Chunyang; Tan, Jindong

doi:10.1109/robio49542.2019.8961677

Cited by 9 publications

(6 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, because data-driven fault diagnosis methods cannot consider the diversity of the programmable motions of a cobot, anomalies may occur in a fixed test program. For example, it was difficult to detect abnormalities in programs designed by an operator [9][10][11][12][13]. Similarly, component-level fault diagnosis models such as motors [7][8], rolling bearings [14][15][16], gears [17][18][19], and sensors [20] are also designed based on predefined test programs.…”

Section: Figure 1 the Complexity Of Programmable Motions In Cobotmentioning

confidence: 99%

“…Therefore, a data-driven approach that analyzes the physical aspects of dynamic scenarios is a promising technique. In [10], a method for detecting faults in gearboxes based on vibration signals from industrial robots was developed. In this method, a health factor that maximizes and quantifies the fault information from vibration signals is defined.…”

Section: Related Workmentioning

confidence: 99%

“…Therefore, it cannot be generalized and applied to all the operation scenarios. Hidden Markov models [10], discrete wavelet transform artificial neural networks [11], and unsupervised learning models using signal analysis [12][13] also support restricted operation environments. Most previous studies performed data-driven failure pattern analysis.…”

Section: Related Workmentioning

confidence: 99%

See 2 more Smart Citations

Programmable Motion-Fault Detection for a Collaborative Robot

2021

View full text Add to dashboard Cite

Smart factories should be able to respond to catastrophic situations proactively, such as recalls caused by production line disruptions and equipment failures. Therefore, the necessity for predictive maintenance technology, such as fault detection or diagnosis of equipment has increased in recent years. In particular, predicting the faults of collaborative robots is becoming increasingly crucial because smart factories pursue efficient collaboration between humans and devices. However, collaborative robots have the characteristic of executing programmable motions designed by an operator, rather than performing fixed tasks. If existing fault diagnosis methods are applied to non-fixed programmable motions, problems arise in terms of setting absolute criteria for fault analysis, interpreting the meanings of detected values, and fault tracking or fault cause analysis. Therefore, we propose a method of programmable motion-fault detection by analyzing motion residuals to solve the three problems mentioned above. The proposed method can expand the fault diagnostic range of collaborative robots.

show abstract

Section: Figure 1 the Complexity Of Programmable Motions In Cobotmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Programmable Motion-Fault Detection for a Collaborative Robot

2021

View full text Add to dashboard Cite

show abstract

“…For example, in [ 5 ], a methodology was proposed to diagnose failures in the ballscrew of a robot, using time-frequency transform methods such as short-time Fourier transform (STFT), and wavelet packet transform (WPT), statistical indicators such as energy, root mean square (RMS), and kurtosis, and to perform a comparison between the classification methods: convolutional neural networks (CNN), logistic regression (LR), and k-nearest neighbors (KNN). Identifying the failure in a rotate vector reducer (RV), by analyzing acoustic emissions (AE), was presented in [ 6 ]. For this, they used WPT, to denoise the acoustic signals, and used a hidden Markov model (HMM), to infer the failure state of the RV.…”

Section: Introductionmentioning

confidence: 99%

Supervised Machine-Learning Methodology for Industrial Robot Positional Health Using Artificial Neural Networks, Discrete Wavelet Transform, and Nonlinear Indicators

Galan-Uribe

Amezquita-Sanchez

Morales-Velazquez

2023

Sensors

View full text Add to dashboard Cite

Robotic systems are a fundamental part of modern industrial development. In this regard, they are required for long periods, in repetitive processes that must comply with strict tolerance ranges. Hence, the positional accuracy of the robots is critical, since degradation of this can represent a considerable loss of resources. In recent years, prognosis and health management (PHM) methodologies, based on machine and deep learning, have been applied to robots, in order to diagnose and detect faults and identify the degradation of robot positional accuracy, using external measurement systems, such as lasers and cameras; however, their implementation is complex in industrial environments. In this respect, this paper proposes a method based on discrete wavelet transform, nonlinear indices, principal component analysis, and artificial neural networks, in order to detect a positional deviation in robot joints, by analyzing the currents of the actuators. The results show that the proposed methodology allows classification of the robot positional degradation with an accuracy of 100%, using its current signals. The early detection of robot positional degradation, allows the implementation of PHM strategies on time, and prevents losses in manufacturing processes.

show abstract

“…Some fault detection methods for the RV reducer based on machine learning, acoustic emission, and artificial neural network technologies are reported in the literature. For example, an acoustic emission-based hidden Markov model is proposed in Zhang et al 7 A machine learning-based method is proposed in Raouf et al 8 A one-shot learning graph neural network method is proposed in Yang et al 9 for the RV reducer. A feature recognition method for the industrial robot RV reducer is proposed in Kim et al 3 An artificial neural network algorithm is introduced in Anand et al, 10 which is used for monitoring the fault of the industrial robot.…”

Section: Introductionmentioning

confidence: 99%

Tooth root crack detection of planet gear in industrial robot RV reducer

Yan

Guo

Liu

2023

Measurement and Control

View full text Add to dashboard Cite

The windowed synchronous averaging (WSA) is widely utilized for planetary structures. However, it cannot be applied for the fault detection of the planetary structure in the industrial robot rotate vector (RV) reducer. The robot usually works within a specified angle range, which causes the RV reducer rotates incompletely. To address this issue, an angle compensation local synchronous fitting scheme is proposed. To detect the localized planet gear fault in the RV reducer, the observed vibration is equi-angle resampled. And the synchronous interference contained in the resampled vibration is constructed and removed according to the angle compensation strategy. The residual data is used to construct the synthetic vibration of the planet gear. Then, the fault feature of the planet gear can be detected. Experiments on the RV reducer test rig under the robot running conditions support the effectiveness of the proposed scheme positively.

show abstract

Industrial Robot Rotate Vector Reducer Fault Detection Based on Hidden Markov Models

Cited by 9 publications

References 27 publications

Programmable Motion-Fault Detection for a Collaborative Robot

Programmable Motion-Fault Detection for a Collaborative Robot

Supervised Machine-Learning Methodology for Industrial Robot Positional Health Using Artificial Neural Networks, Discrete Wavelet Transform, and Nonlinear Indicators

Tooth root crack detection of planet gear in industrial robot RV reducer

Contact Info

Product

Resources

About