Journal of the Computational Structural Engineering Institute of Korea (한국전산구조공학회논문집)

Computational Structural Engineering Institute of Korea (한국전산구조공학회)
권호 표지
DOI QR코드

DOI QR Code

Development of Classification Model on SAC Refrigerant Charge Level Using Clustering-based Steady-state Identification

군집화 기반 정상상태 식별을 활용한 시스템 에어컨의 냉매 충전량 분류 모델 개발

  • Jae-Hee, Kim (School of Mechanical Engineering, Pusan Nat'l University) ;
  • Yoojeong, Noh (School of Mechanical Engineering, Pusan Nat'l University) ;
  • Jong-Hwan, Jeung (SAC Research/Engineering Division, LG Electronics) ;
  • Bong-Soo, Choi (SAC Research/Engineering Division, LG Electronics) ;
  • Seok-Hoon, Jang (SAC Research/Engineering Division, LG Electronics)
  • 김재희 (부산대학교 기계공학부) ;
  • 노유정 (부산대학교 기계공학부) ;
  • 정종환 (LG전자 SAC개발실) ;
  • 최봉수 (LG전자 SAC개발실) ;
  • 장석훈 (LG전자 SAC개발실)
  • Received : 2022.10.18
  • Accepted : 2022.11.15
  • Published : 2022.12.31
Download PDF
⟨ Previous Next ⟩

Abstract

Refrigerant mischarging is one of the most frequently occurring failure modes in air conditioners, and both undercharging and overcharging degrade cooling performance. Therefore, it is important to accurately determine the amount of charged refrigerant. In this study, a support vector machine (SVM) model was developed to multi-classify the refrigerant mischarge through steady-state identification via fuzzy clustering techniques. For steady-state identification, a fuzzy clustering algorithm was applied to the air conditioner operation data using the difference between moving averages. The identification results using the proposed method were compared with those using existing steady-state determination techniques studied through the inversed Fisher's discriminant ratio (IFDR). Subsequently, the main features were selected using minimum redundancy maximum relevance (mRMR) considering the correlation among candidate features, and an SVM multi-classification model was devised using the derived features. The proposed method achieves satisfactory accuracy and robustness from test data collected in the new domain.

냉매 오충전은 에어컨에서 빈번하게 발생하는 고장 모드 중 하나로, 적정 충전량 대비 부족 및 과충전 모두 냉방 성능의 저하를 유발하므로 충전된 냉매량을 정확하게 판단하는 것이 중요하다. 본 연구에서는 퍼지 군집화 기법을 통한 정상상태 식별을 통해 냉매 오충전량을 다중 분류하는 모델을 개발하였다. 정상상태 식별을 위해 에어컨 운전 데이터에 대해 이동 평균 간의 차이를 활용한 퍼지 군집화 알고리즘을 적용하였으며, IFDR를 통해 기존 연구된 정상상태 판단 기법들과 식별 결과를 비교하였다. 이후, 시스템 내 상관성을 고려한 mRMR을 이용해 특징을 선택하였으며, 도출된 특징을 이용해 SVM 기반의 다중 분류 모델이 생성되었다. 제안된 방법은 시험 데이터를 통해 만족할 만한 분류 정확도와 강건성을 도출하였다.

Keywords

Acknowledgement

이 논문은 부산대학교 기본연구지원사업(2년)에 의하여 연구되었음.

References

  1. Cao, S., Rhinehart, R.R. (1995) An Efficient Method for On-line Identification of Steady State, J. Process Control, 5(6), pp. 363~374. https://doi.org/10.1016/0959-1524(95)00009-F
  2. Cortes, C., Vapnik, V. (1995) Support-vector Networks, Mach. Learn., 20(3), pp.273~297.
  3. Dunn, J. C. (1973) A Fuzzy Relative of the ISODATA Process and Its Use in Detecting Compact Well-Separated Clusters, J. Cybern., 3(3), pp.32~57. https://doi.org/10.1080/01969727308546046
  4. Kim, M., Yoon, S.H., Domanski, P.A., Payne, W.V. (2008) Design of a Steady-state Detector for Fault Detection and Diagnosis of a Residential Air Conditioner, Int. J. Refrig., 31(5), pp.790~799. https://doi.org/10.1016/j.ijrefrig.2007.11.008
  5. Kim, S., Kang, Y.-J., Noh, Y., Park, S., Ahn, B. (2021) Fault Classification Model based on Time Domain Feature Extraction of Vibration Data, J. Comput. Struct. Eng. Inst. Korea, 34(1), pp.25~33. https://doi.org/10.7734/COSEIK.2021.34.1.25
  6. Kim, S., Noh, Y., Kang, Y.-J., Park, S., Lee, J.W., Chin, S.W. (2022) Hybrid Data-scaling Method for Fault Classification of Compressors, Meas., 201, p.111619.
  7. Kim, W., Braun, J.E. (2012) Evaluation of the Impacts of Refrigerant Charge on Air Conditioner and Heat Pump Performance, Int. J. Refrig., 35, pp.1805~1814. https://doi.org/10.1016/j.ijrefrig.2012.06.007
  8. Li, Z., Shi, S., Chen, H., Wei, W., Wang, Y., Liu, Q., Liu, T. (2020) Machine Learning based Diagnosis Strategy for Refrigerant Charge Amount Malfunction of Variable Refrigerant Flow System, Int. J. Refrig., 110, pp.95~105. https://doi.org/10.1016/j.ijrefrig.2019.10.026
  9. Liu, A.J., Mukherjee, A., Hu, L., Chen, J., Nair, V.N. (2022) Performance and Interpretability Comparisons of Supervised Machine Learning Algorithms: An Empirical Study, arXiv preprint arXiv :2204.12868.
  10. Peng, H., Long, F., Ding, C. (2005) Feature Selection based on Mutual Information Criteria of Max-Dependency, Max-Relevance, and Min-Redundancy, IEEE Trans. Pattern Anal. & Mach. Intell., 27, pp.1226~1238. https://doi.org/10.1109/TPAMI.2005.159
  11. Sun, K., Li, G., Chen, H., Liu, J., Li, J., Hu, W. (2016) A novel Efficient SVM-based Fault Diagnosis Method for Multi-Split Air Conditioning System's Refrigerant Charge Fault Amount, Appl. Therm. Eng., 108, pp.989~998. https://doi.org/10.1016/j.applthermaleng.2016.07.109
  12. Wan, H., Cao, T., Hwang, Y., Chang, S.D., Yoon, Y.J. (2021) Machine-Learning-based Compressor Models: A Case Study for Variable Refrigerant Flow Systems, Int. J. Refrig., 123, pp.23~33. https://doi.org/10.1016/j.ijrefrig.2020.12.003
  13. Webb, A.R. (2003) Statistical Pattern Recognition. John Wiley & Sons.
  14. Xu, D., Tian, Y. (2015) A Comprehensive Survey of Clustering Algorithms, Ann. Data Sci., 2, pp.165~193. https://doi.org/10.1007/s40745-015-0040-1
  15. Yoo, J.W., Hong, S.B., Kim, M.S. (2017) Refrigerant Leakage Detection in an EEV Installed Residential Air Conditioner with Limited Sensor Installations, Int. J. Refrig., 78, pp. 157~165. https://doi.org/10.1016/j.ijrefrig.2017.03.001