All Issue

2025 Vol.3, Issue 2 Preview Page

Research Article

Comparative study for enhancing PCA-Based Anomaly Detection with Missing Values

결측치 처리에 따른 PCA 기반 구조물 이상상태 탐지 신뢰성 향상 연구

Jaeseok Jung1
,
Seung-Seop Jin2*
정 재석1
,
진 승섭2*
1Integrated Bachelor’s and Master’s Program Student, Department of Civil and Environmental Engineering, Sejong University, Seoul 05006, Korea
2Assistant Professor, Department of Civil and Environmental Engineering, Sejong University, Seoul 0556, Korea
1세종대학교 건설환경공학과 학석사연계과정
2세종대학교 건설환경공학과 조교수
*Corresponding Author
31 December 2025. pp. 123-136
PDF XML Citation
Abstract

Vibration-based Structural Health Monitoring (SHM) provides modal information, such as natural frequencies, which is essential for structural state diagnosis. However, depending on external factors like excitation conditions, certain modes are frequently missing over time. These frequent missing values are a primary factor that degrades anomaly detection performance. In this study, a fundamental investigation was conducted to enhance the reliability of Principal Component Analysis (PCA)-based structural anomaly detection in realistic environments containing missing values. To achieve this, we compared imputation-based approaches (KNN, MICE), which fill in missing values prior to PCA, with a method that intrinsically handles missing values within the PCA algorithm itself. Numerical experiments considering various missing rates and patterns were conducted to evaluate the performance of these methodologies, and based on the results, directions for future research are proposed.

Keywords
Vibration-based Structural Health Monitoring
Anomaly Detection
Missing Values
Principal Component Analysis

진동 기반 구조물 건전성 모니터링(SHM)은 구조물 상태 진단에 필요한 모드 정보(고유진동수 등)를 통해 제공한다. 하지만 가진 조건 등 외부 조건에 따라 일부 모드들이 시간에 따라 누락되어 결측되는 문제가 존재한다. 이러한 잦은 결측치는 탐지 성능을 저해하는 주요 원인 중 하나이다. 본 연구에서는 결측치가 발생하는 실제 환경에서도 주성분 분석(PCA) 기반 구조물 이상상태 탐지의 신뢰성을 향상시키는 기초 연구를 수행하였다. 이를 위해 PCA 적용 전 결측치를 보간하는 방법(KNN, MICE)과 PCA 알고리즘 자체 내에서 결측치를 처리할 수 있는 방법을 비교하였다. 다양한 결측률 및 결측치 유형에 대한 수치 실험을 상기 방법론들의 성능을 평가하였으며, 그 결과를 토대로 향후 연구 방향을 제시하였다.

키워드
진동기반 구조물 건전성 모니터링
이상상태 탐지
결측치
주성분 분석
References
1

Cho, B., Dayrit, T., Gao, Y., Wang, Z., Hong, T., Sim, A. and Wu, K. (2020). Effective missing value imputation methods for building monitoring data. 2020 IEEE International Conference on Big Data (Big Data), 2866-2875.

10.1109/BigData50022.2020.9378230
2

Cormen, T. H., Leiserson, C. E., Rivest, R. L. and Stein, C. (2009). Introduction to Algorithms (3rd ed.). MIT Press.

3

Doebling, S. W., Farrar, C. R., Prime, M. B. and Shevitz, D. W. (1996). Damage identification and health monitoring of structural and mechanical systems from changes in their vibration characteristics: A literature review. LA-13070-MS, Los Alamos National Laboratory, New Mexico, USA.

10.2172/249299
4

Jackson, J. E. and Mudholkar, G. S. (1979). Control procedures for residuals associated with principal component analysis. Technometrics 21(3), 341-349

10.1080/00401706.1979.10489779
5

Jin, S. S. and Jung, H. J. (2018). Vibration-based damage detection using online learning algorithm for output-only structural health monitoring. Structural Health Monitoring 17(4), 727-746.

10.1177/1475921717717310
6

Jin, S.-S., Cho, S. and Jung, H.-J. (2015). Adaptive reference updating for vibration-based structural health monitoring under varying environmental conditions. Computers & Structures 158, 211-224.

10.1016/j.compstruc.2015.06.001
7

Lei, X. and Sun, L. (2020). Lost data reconstruction for structural health monitoring using deep convolutional generative adversarial networks. Structural Health Monitoring 20(4), 2069-2087.

10.1177/1475921720959226
8

Pedregosa, F., Varoquaux, G., Gramfort, A., Michel, V., Thirion, B., Grisel, O., Blondel, M., Prettenhofer, P., Weiss, R., Dubourg, V., Vanderplas, J., Passos, A., Cournapeau, D., Brucher, M., Perrot, M. and Duchesnay, É. (2011). Scikit-learn: Machine Learning in Python. Journal of Machine Learning Research 12, 2825-2830.

9

Tipping, M. E. and Bishop, C. M. (1999). Probabilistic Principal Component Analysis. Journal of the Royal Statistical Society Series B: Statistical Methodology 61(3), 611-622.

10.1111/1467-9868.00196
10

Yan, A. M., Kerschen, G., De Boe, P. and Golinval, J.-C. (2005). Structural damage diagnosis under varying environmental conditions – Part I: A linear analysis. Mechanical Systems and Signal Processing 19, 847-864.

10.1016/j.ymssp.2004.12.002
11

Zhang, R., Chen, Z., Chen, S., Zheng, J., Buyukozturk, O. and Sun, H. (2019). Deep long short-term memory networks for nonlinear structural seismic response prediction. Computers & Structures 220, 55-68.

10.1016/j.compstruc.2019.05.006
Information
  • Publisher :Korean Institute of Bridge and Structural Engineers
  • Publisher(Ko) :한국교량및구조공학회
  • Journal Title :Journal of Structure Research and Practice
  • Journal Title(Ko) :한국교량및구조공학회 논문집
  • Volume : 3
  • No :2
  • Pages :123-136
  • Received Date : 2025-11-03
  • Revised Date : 2026-01-04
  • Accepted Date : 2026-01-05
  • DOI :https://doi.org/10.22725/JSRP.2025.3.2.123
Journal Informaiton Journal of Structure Research and Practice Journal of Structure Research and Practice
Online Submission & Review System
  • KOFST
  • KISTI Current Status
  • KISTI Cited-by
  • KCI 문헌 유사도 검사
  • orcid
  • open access
  • ccl
Journal Informaiton Journal Informaiton - close