기계장비 진동 데이터를 이용한 딥러닝 기반의 고장 분류 모델
Fault Classification Model Based on Deep Learning Using Vibration Data of Mechanical Equipment
0
조회
0
다운로드
4
인용
0
참고문헌
키워드
초록
최근 4차 산업 혁명으로 인해 공장에서는 기계 시설물의 고장으로 인한 제조 및 생산 시간 증가와 수리비용 증가를 최신 기술을 통해 예방하고자 한다. 그리하여 연구자들은 안전사고 및 고장을 예지(豫知)하고, 제품 불량 등으로 인한 시민 불편, 사회적 혼란 등의 문제를 방지하기 위해 노력하고 있다. 이런 문제 해결을 위해 장비에 부착된 센서로부터 받은 데이터를 통해 기계 상황을 모니터링 및 예측 가능한 시스템이 요구된다. 본 논문에서는 기계설비에서 생성되는 시계열 데이터를 다양한 딥러닝 기반의 시계열 분류 모델에서 비교분석을 수행하였다. 총 13개의 모델을 사용하여 어떤 딥러닝 모델이 학습에 효율적이고 성능이 좋은지 실험을 수행하였고, 시계열의 복잡한 패턴과 시간 및 공간 패턴을 효과적으로 학습하는 CNN 계열 모델이 정확도, 정밀도, 재현율, F1 성능지표에서 100% 성능을 달성하여 우수한 것을 확인할 수 있었다. 기존까지 진동 시계열 데이터에 대해 다양한 딥러닝 모델의 비교분석이 연구는 없었기에 본 논문의 결과를 통해 다른 기계 시설물의 고장 분류에 있어서 도움이 될 것이라 예상된다.
Abstract
Due to the recent Fourth Industrial Revolution, factories want to prevent the increase in manufacturing and production time and repair costs due to the failure of mechanical facilities through the latest technology. Therefore, researchers are trying to predict safety accidents and failures, and to prevent problems such as civil inconvenience and social confusion caused by product defects. To solve this problem, a system that can monitor and predict the machine situation through data received from sensors attached to the equipment is required. In this paper, comparative analysis was performed on time series data generated in machine facilities in various deep learning-based time series classification models. A total of 13 models were used to experiment with which deep learning models were efficient and performing well, and the models that effectively learned time and space patterns of time series recorded 100% performance in accuracy, precision, reproducibility, and F1 performance indicators. Since there have been no studies on the comparative analysis of various deep learning models for vibration time series data until now, the results of this paper are expected to help in classifying failures in other mechanical facilities.
참고문헌 (20)
- "기계시설물 고장 예지 센서 소개" (2021)[URL]
- Wang, Zhiguang, "Time series classification from scratch with deep neural networks: A strong baseline", 2017International joint conference on neural networks (IJCNN) (2017), IEEE
- Ismail Fawaz, Hassan, "Inceptiontime:Finding alexnet for time series classification", Data Mining and Knowledge Discovery, 34(6), pp.1936~1962 (2020)
- Zou, Xiaowu, "Integration of residual network and convolutional neural network along with various activation functions and global pooling for time series classification", Neurocomputing, 367(2019), pp.39~45 (2019)
- Fauvel, Kevin, "XCM: An Explainable Convolutional Neural Network for Multivariate Time Series Classification", Mathematics, 9(23), pp.3137~ (2021)
- Dempster, Angus, "Minirocket: A very fast(almost) deterministic transform for time series classification", Proceedings of the 27th ACM SIGKDD Conference on Knowledge Discovery & Data Mining (2021)
- Dempster, Angus, "ROCKET: exceptionally fa st and accurate time series classification using random convolutional kernels", Data Mining and Knowledge Discovery, 34(5), pp.1454~1495 (2020)
- Hochreiter, Sepp, "Long short-term memory", Neural computation, 9(8), pp.1735~1780 (1997)
- Karim, Fazle, "LSTM fully convolutional networks for time series classification", IEEE access, 6, pp.1662~1669 (2017)
- Elsayed, Nelly, "Gated recurrent neural networks empirical utilization for time series classification", 2019 International Conference on Internet of Things(iThings) and IEEE Green Computing and Communications(GreenCom) and IEEE Cyber, Physical and Social Computing(CPSCom)and IEEE Smart Data(SmartData) (2019), IEEE
- Ismail Fawaz, Hassan, "Deep learning for time series classification: a review", Data mining and knowledge discovery, 33(4), pp.917~963 (2019)
- Karim, Fazle, "Multivariate LSTM-FCNs for time series classification", Neural Networks, 116, pp.237~245 (2019)
- Fawaz, Hassan Ismail, "Deep neural network ensembles for time series classification", 2019International Joint Conference on Neural Networks(IJCNN) (2019), IEEE
- Karim, Fazle, "Insights into LSTM fully convolutional networks for time series classification", IEEE Access, 7, pp.67718~67725 (2019)
- Fawaz, Hassan Ismail, "Transfer learning for time series classification", 2018 IEEE international conference on big data(Big Data) (2018), IEEE