KIPS Transactions on Software and Data Engineering (정보처리학회논문지:소프트웨어 및 데이터공학)

Korea Information Processing Society (한국정보처리학회)
권호 표지
DOI QR코드

DOI QR Code

Learning Method for Regression Model by Analysis of Relationship Between Input and Output Data with Periodicity

주기성을 갖는 입출력 데이터의 연관성 분석을 통한 회귀 모델 학습 방법

  • 김혜진 (아주대학교 전자공학과) ;
  • 박예슬 (아주대학교 AI융합네트워크학과) ;
  • 이정원 (아주대학교 전자공학과/AI융합네트워크학과)
  • Received : 2021.11.23
  • Accepted : 2022.03.08
  • Published : 2022.07.31
Download PDF
⟨ Previous Next ⟩

Abstract

In recent, sensors embedded in robots, equipment, and circuits have become common, and research for diagnosing device failures by learning measured sensor data is being actively conducted. This failure diagnosis study is divided into a classification model for predicting failure situations or types and a regression model for numerically predicting failure conditions. In the case of a classification model, it simply checks the presence or absence of a failure or defect (Class), whereas a regression model has a higher learning difficulty because it has to predict one value among countless numbers. So, the reason that regression modeling is more difficult is that there are many irregular situations in which it is difficult to determine one output from a similar input when predicting by matching input and output. Therefore, in this paper, we focus on input and output data with periodicity, analyze the input/output relationship, and secure regularity between input and output data by performing sliding window-based input data patterning. In order to apply the proposed method, in this study, current and temperature data with periodicity were collected from MMC(Modular Multilevel Converter) circuit system and learning was carried out using ANN. As a result of the experiment, it was confirmed that when a window of 2% or more of one cycle was applied, performance of 97% or more of fit could be secured.

최근 로봇이나 설비, 회로 등에 센서 내장이 보편화 되고, 측정된 센서 데이터를 학습하여 기기의 고장을 진단하기 위한 연구가 활발하게 수행되고 있다. 이러한 고장 진단 연구는 고장 상황이나 종류를 예측하기 위한 분류(Classification) 모델 개발과 정량적으로 고장 상황을 예측하기 위한 회귀(Regression) 모델 개발로 구분된다. 분류 모델의 경우, 단순히 고장이나 결함의 유무(Class)를 확인하는 반면, 회귀 모델은 무수히 많은 수치 중에 하나의 값(Value)을 예측해야 하므로 학습 난이도가 더 높다. 즉, 입력과 출력을 대응시켜 고장을 예측을 할 때, 유사한 입력값이 동일한 출력을 낸다고 결정하기 어려운 불규칙한 상황이 다수 존재하기 때문이다. 따라서 본 논문에서는 주기성을 지닌 입출력 데이터에 초점을 맞추어, 입출력 관계를 분석하고, 슬라이딩 윈도우 기반으로 입력 데이터를 패턴화 하여 입출력 데이터 간의 규칙성을 확보하도록 한다. 제안하는 방법을 적용하기 위해, 본 연구에서는 MMC(Modular Multilevel Converter) 회로 시스템으로부터 주기성을 지닌 전류, 온도 데이터를 수집하여 ANN을 이용하여 학습을 진행하였다. 실험 결과, 한 주기의 2% 이상의 윈도우를 적용하였을 때, 적합도 97% 이상의 성능이 확보될 수 있음을 확인하였다.

Keywords

Acknowledgement

이 논문은 정부(과학기술정보통신부)의 재원으로 한국연구재단의 지원을 받아 수행된 연구임(No. 2020R1A2C1007400).

References

  1. Y. Liu, X. Yan, C. Zhang, and W. Liu, "An ensemble convolutional neural networks for bearing fault diagnosis using multi-sensor data," Sensors, Vol.19, No.23, pp.5300, 2019. https://doi.org/10.3390/s19235300
  2. Z. Li, R. Outbib, S. Giurgea, D. Hissel, A. Giraud, and P. Couderc, "Fault diagnosis for fuel cell systems: A data-driven approach using high-precise voltage sensors," Renewable Energy, Vol.135, pp.1435-1444, 2019. https://doi.org/10.1016/j.renene.2018.09.077
  3. S. S. Udmale, S. K. Singh, R. Singh, and A. K. Sangaiah, "Multi-fault bearing classification using sensors and ConvNet-based transfer learning approach," IEEE Sensors Journal, Vol.20, No.3, pp.1433-1444, 2019. https://doi.org/10.1109/jsen.2019.2947026
  4. D. A. T. Tran, Y. Chen, and C. Jiang, "Comparative investigations on reference models for fault detection and diagnosis in centrifugal chiller systems," Energy and Buildings, Vol.133, pp.246-256, 2016. https://doi.org/10.1016/j.enbuild.2016.09.062
  5. M. S. Mirnaghi and F. Haghighat, "Fault detection and diagnosis of large-scale HVAC systems in buildings using data-driven methods: A comprehensive review," Energy and Buildings, Vol.229, pp.110492, 2020. https://doi.org/10.1016/j.enbuild.2020.110492
  6. A. Capozzoli, F. Lauro, and I. Khan, "Fault detection analysis using data mining techniques for a cluster of smart office buildings," Expert Systems with Applications, Vol.42, No.9, pp.4324-4338, 2015. https://doi.org/10.1016/j.eswa.2015.01.010
  7. J. A. Cortes-Ibanez, S. Gonzalez, J. J. Valle-Alonso, J. Luengo, S. Garcia, and F. Herrera, "Preprocessing methodology for time series: an industrial world application case study," Information Sciences, Vol.514, pp.385-401, 2020. https://doi.org/10.1016/j.ins.2019.11.027
  8. H. Kaneko and K. Funatsu, "Fast optimization of hyperparameters for support vector regression models with highly predictive ability," Chemometrics and Intelligent Laboratory Systems, Vol.142, pp.64-69, 2015. https://doi.org/10.1016/j.chemolab.2015.01.001
  9. A. Ahrens, C. B. Hansen, and M. E. Schaffer, "lassopack: Model selection and prediction with regularized regression in Stata," The Stata Journal, Vol.20, No.1, pp.176-235, 2020. https://doi.org/10.1177/1536867x20909697
  10. S. Selvin, R. Vinayakumar, E. A. Gopalakrishnan, V. K. Menon, & K. P. Soman, "Stock price prediction using LSTM, RNN and CNN-sliding window model," 2017 international conference on advances in computing, communications and informatics (icacci). IEEE, 2017.
  11. H. S. Hota, R. Handa, and A. K. Shrivas, "Time series data prediction using sliding window based RBF neural network," International Journal of Computational Intelligence Research, Vol.13, No.5, pp.1145-1156, 2017.
  12. S. Park, S. Jung, S. Jung, S. Rho, and E. Hwang, "Sliding window-based LightGBM model for electric load forecasting using anomaly repair," The Journal of Supercomputing, Vol.77, No.11, pp.12857-12878, 2021. https://doi.org/10.1007/s11227-021-03787-4
  13. O. Renaud and M. P. Victoria-Feser, "A robust coefficient of determination for regression," Journal of Statistical Planning and Inference, Vol.140, No.7, pp.1852-1862, 2010. https://doi.org/10.1016/j.jspi.2010.01.008