항공엔진 열화데이터 기반 잔여수명 예측력 향상을 위한 데이터 전처리 방법 연구
A study on Data Preprocessing for Developing Remaining Useful Life Predictions based on Stochastic Degradation Models Using Air Craft Engine Data
2
조회
0
다운로드
3
인용
0
참고문헌
키워드
초록
Recently, a study of prognosis and health management (PHM) was conducted to diagnose failure and predict the life of air craft engine parts using sensor data. PHM is a framework that provides individualized solutions for managing system health. This study predicted the remaining useful life (RUL) of aeroengine using degradation data collected by sensors provided by the IEEE 2008 PHM Conference Challenge. There are 218 engine sensor data that has initial wear and production deviations. It was difficult to determine the characteristics of the engine parts since the system and domain-specific information was not provided. Each engine has a different cycle, making it difficult to use time series models. Therefore, this analysis was performed using machine learning algorithms rather than statistical time series models. The machine learning algorithms used were a random forest, gradient boost tree analysis and XG boost. A sliding window was applied to develop RUL predictions. We compared model performance before and after applying the sliding window, and proposed a data preprocessing method to develop RUL predictions. The model was evaluated by R-square scores and root mean squares error (RMSE). It was shown that the XG boost model of the random split method using the sliding window preprocessing approach has the best predictive performance.
Abstract
Recently, a study of prognosis and health management (PHM) was conducted to diagnose failure and predict the life of air craft engine parts using sensor data. PHM is a framework that provides individualized solutions for managing system health. This study predicted the remaining useful life (RUL) of aeroengine using degradation data collected by sensors provided by the IEEE 2008 PHM Conference Challenge. There are 218 engine sensor data that has initial wear and production deviations. It was difficult to determine the characteristics of the engine parts since the system and domain-specific information was not provided. Each engine has a different cycle, making it difficult to use time series models. Therefore, this analysis was performed using machine learning algorithms rather than statistical time series models. The machine learning algorithms used were a random forest, gradient boost tree analysis and XG boost. A sliding window was applied to develop RUL predictions. We compared model performance before and after applying the sliding window, and proposed a data preprocessing method to develop RUL predictions. The model was evaluated by R-square scores and root mean squares error (RMSE). It was shown that the XG boost model of the random split method using the sliding window preprocessing approach has the best predictive performance.
참고문헌 (21)
- 안건이, "다중 센서 데이터 분석을 위한 설명 가능한 합성곱신경망 모델", 대한산업공학회지, 45(2), pp.146~153 (2019), 대한산업공학회[DOI]
- 백재욱, "고장예지 및 건전성관리에서의 AI 기법", 신뢰성 응용연구, 19(3), pp.243~255 (2019), 한국신뢰성학회[DOI]
- Bontempi, G., "European Business Intelligence Summer School", pp.62~77 (2012), Springer
- Breiman, L., "Random forests", Machine Learning, 45(1), pp.5~32 (2001)
- 장중혁, "데이터 스트림 마이닝에서 정보 중요성 차별화를 위한 퍼지 윈도우 기법", 한국산학기술학회논문지, 12(9), pp.4183~4191 (2011), 한국산학기술학회
- Chen, T., "Xgboost : A scalable tree boosting system", Proceedings of the 22nd acm Sigkdd International Conference on Knowledge Discovery and Datamining, pp.785~794 (2016)
- Chiu, S. C., "Incremental mining of closed inter-transaction itemsets over data stream sliding window", Journal of Information Science, 37(2), pp.208~220 (2011)
- 최우성, "머신러닝 기반 고온 부품 열화 평가 프로그램 개발", 대한기계학회논문집 A, 44(1), pp.57~62 (2020), 대한기계학회
- Coble, J.B., "Prognostic algorithm categorization with PHM challenge application", 2008 International Conference on Prognostics and Health Management, pp.1~6 (2008), IEEE
- Friedman, J.H., "Greedy function approximation : a gradient boosting machine", Annals of Statistics, pp.1189~1232 (2001)
- Heimes, F.O., "Recurrent Neural Networks for Remaining Useful Life Estimation", 2008 International Conference on Prognostics and Health Management, pp.1~6 (2008), IEEE
- 정훈, "기계적 모터 고장진단을 위한 머신러닝 기법", 한국산업경영시스템학회지, 40(1), pp.57~64 (2017), 한국산업경영시스템학회
- Koc, C. K., "Analysis of sliding window techniques for exponentiation", Computers and Mathematics with Applications, 30(10), pp.17~24 (1995)
- Laguna, J.O., "A dynamic sliding window approach for activity recognition", International Conference on User Modeling, Adaptation, and Personalization, pp.219~230 (2011), pringer
- Le Son, K., "Remaining useful life estimation based on stochastic deterioration models : A comparative study", Reliability Engineering and System Safety, 112, pp.165~175 (2013)
- Lee, H.S., "A Study on the development of time series forecasting model for corporate credit risk using machine learning", The Koran Society of Management information Systems, pp.396~405 (2019)
- Peel, L., "Data driven prognostics using a kalman filter ensemble of neural network models", 2008 International Conference on Prognostics And Health Management, pp.1~6 (2008), IEEE
- 서성보, "슬라이딩 윈도우 기반 다변량 스트림 데이타 분류 기법", 정보과학회논문지 : 데이타베이스, 33(2), pp.163~174 (2006), 한국정보과학회
- 심현수, "고장 예지 및 건전성 관리에 관한 문헌연구: 2013~2018", 신뢰성 응용연구, 19(1), pp.68~84 (2019), 한국신뢰성학회[DOI]
- 송현석, "승용차량용 쇽업소버의 열화분석 사례연구", 신뢰성 응용연구, 17(3), pp.181~187 (2017), 한국신뢰성학회
- Wang, T., "A similaritybased prognostics approach for remaining useful life estimation of engineered systems", 2008 International Conference on Prognostics and Health Management, pp.1~6 (2008), IEEE