설비 결함 식별 최적화를 위한 오토인코더 기반N 분할 주파수 영역 이상 탐지
Autoencoder Based N-Segmentation Frequency Domain Anomaly Detection for Optimization of Facility Defect Identification
1
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
제조 분야 설비 예지보전을 위해서 진동, 전류, 온도 등 물리 데이터를 기반으로 설비 이상을 탐지하는 인공지능 학습 모델이 활용되고 있다. 설비 결함, 고장 등 설비 이상 유형은 매우 다양하므로, 주로 오토인코더 기반 비지도 학습 모델을 이용한 이상 탐지 방법이 적용되고 있다. 설비상태의 정상, 비정상 여부는 오토인코더의 재구성 오차를 이용해 효과적으로 분류할 수 있지만, 설비 이상의 구체적인 상태를 식별하는 데 한계가있다. 설비 불균형, 정렬 불량, 고정 불량 등 설비 이상 상황 발생 시, 설비 진동 주파수는 특정 영역에서 정상 상태와 다른 패턴을 나타낸다. 본 논문에서는 전체 진동 주파수 범위를 N개 영역으로 나누어 이상 탐지를 수행하는 N 분할 이상 탐지 방법을 제시하였다. 압축기의 진동 데이터를이용해 주파수와 강도를 달리한 9종의 이상 데이터를 대상으로 실험한 결과, N 분할을 적용하였을 때 더 높은 이상 탐지 성능을 나타냈다. 제안방법은 설비 이상 탐지 이후, 설비 이상 구체화에 활용될 수 있다
Abstract
Artificial intelligence models are being used to detect facility anomalies using physics data such as vibration, current, and temperaturefor predictive maintenance in the manufacturing industry. Since the types of facility anomalies, such as facility defects and failures, anomalydetection methods using autoencoder-based unsupervised learning models have been mainly applied. Normal or abnormal facilityconditions can be effectively classified using the reconstruction error of the autoencoder, but there is a limit to identifying facility anomaliesspecifically. When facility anomalies such as unbalance, misalignment, and looseness occur, the facility vibration frequency shows a patterndifferent from the normal state in a specific frequency range. This paper presents an N-segmentation anomaly detection method thatperforms anomaly detection by dividing the entire vibration frequency range into N regions. Experiments on nine kinds of anomaly datawith different frequencies and amplitudes using vibration data from a compressor showed better performance when N-segmentation wasapplied. The proposed method helps materialize them after detecting facility anomalies.
참고문헌 (25)
- Z. M. Çınar, "Machine learning in predictive maintenance towards sustainable smart manufacturing in industry 4. 0", Sustainability, 19(19), pp.2020~ (2020)[DOI]
- S. Namuduri, "Review-deep learning methods for sensor based predictive maintenance and future perspectives for electrochemical sensors", Journal of The Electrochemical Society, 167(3) (2020)[DOI]
- K. Park, "Anomaly detection in combined driving system based on unsupervised learning", Journal of the Korean Society for Precision Engineering, 40(11) (2023)[DOI]
- Y. K. Lee, "Real-time defect monitoring of laser micro-drilling using reflective light and machine learning models", International Journal of Precision Engineering and Manufacturing, 25 (2024)[DOI]
- M. Pech, "Predictive main"tenance and intelligent sensors in smart factory", Sensors, 21(4) (2021)[DOI]
- N. Azeem, "Experimental study on the Condition Monitoring of Shaft Unbalance by using Vibrations Spectrum and phase Analysis", Proceedings of the Condition Monitoring and Diagnosis (2018)
- N. Azeem, "Experimental condition monitoring for the detection of misaligned and cracked shafts by order analysis", Advances in Mechanical Engineering, 11(5) (2019)[DOI]
- D. Mohamed, "Detection of unbalance and looseness faults in a ventilation turbine using vibration signature analysis", IOP Conference Series : Materials Science and Engineering, 1244(1), pp.2022~ (2022)[DOI]
- S. A. Mohammed, "Fault diagnosis of crack on gearbox using vibration-based approaches", Symmetry, 14 (2022)[DOI]
- R. E. Bouisfi, "Supervised classification of induction motors faults", Proceedings of the International Conference on Energy and Green Computing (2021)
- T. Chen, "Unsupervised anomaly detection of industrial robots using slidingwindow convolutional variational autoencoder", IEEE Access, 8, pp.47072~47081 (2020)[DOI]
- S. Givnan, "Anomaly detection using autoencoder reconstruction upon industrial motors", Sensors, 22 (2022)[DOI]
- Y. Huang, "Motor fault detection and feature extraction using RNN-Based variational autoencoder", IEEE Access, 7, pp.139086~139096 (2019)[DOI]
- T. Tziolas, "Autoencoders for anomaly detection in an industrial multivariate time series dataset", Engineering Proceedings, 18, pp.2022~ (2022)[DOI]
- I. Thoidis, "Semisupervised machine condition monitoring by learning deep discriminative audio features", Electronics, 10 (2021)[DOI]
- Ö. Yilmaz, "Misalignment fault detection by wavelet analysis of vibration signals", International Advanced Researches and Engineering Journal, 3(3), pp.156~163 (2019)[DOI]
- B. Cahyono, "Vibration spectrum analysis for indicating damage on turbine and steam generator amurang unit 1", International Journal of Marine Engineering Innovation and Research, 2 (2017)[DOI]
- S. Ahmad, "Autoencoder-based condition monitoring and anomaly detection method for rotating machines", Proceedings of IEEE International Conference on Big Data (2020)
- I. Ahmed, "A Smart Anomaly-Detection System for Industrial Machines Based on Feature Autoencoder and Deep Learning", Micromachines, 14 (2023)[DOI]
- M. Rao, "A speed normalized autoencoder for rotating machinery fault detection under varying speed conditions", Mechanical Systems and Signal Processing, 189 (2023)[DOI]
- O. Serradilla, "Adaptable and explainable predictive maintenance : Semi-supervised deep learning for anomaly detection and diagnosis in press machine data", Applied Sciences, 11 (2021)[DOI]
- T. Tziolas, "Autoencoders for anomaly detection in an industrial multivariate time series dataset", Engineering Proceedings, 18 (2022)[DOI]