Journal of Korea Water Resources Association (한국수자원학회논문집)

Korea Water Resources Association (한국수자원학회)
권호 표지
DOI QR코드

DOI QR Code

Comparison of ANN model's prediction performance according to the level of data uncertainty in water distribution network

상수도관망 내 데이터 불확실성에 따른 절점 압력 예측 ANN 모델 수행 성능 비교

  • Jang, Hyewoon (School of Civil, Environmental and Architectural Engineering, Korea University) ;
  • Jung, Donghwi (School of Civil, Environmental and Architectural Engineering, Korea University) ;
  • Jun, Sanghoon (Future and Fusion Lab of Civil, Environmental and Architectural Engineering, Korea University)
  • 장혜운 (고려대학교 건축사회환경공학부) ;
  • 정동휘 (고려대학교 건축사회환경공학부) ;
  • 전상훈 (고려대학교 미래건설환경융합연구소)
  • Received : 2022.11.04
  • Accepted : 2022.12.01
  • Published : 2022.12.31
Download PDF
⟨ Previous Next ⟩

Abstract

As the role of water distribution networks (WDNs) becomes more important, identifying abnormal events (e.g., pipe burst) rapidly and accurately is required. Since existing approaches such as field equipment-based detection methods have several limitations, model-based methods (e.g., machine learning based detection model) that identify abnormal events using hydraulic simulation models have been developed. However, no previous work has examined the impact of data uncertainties on the results. Thus, this study compares the effects of measurement error-induced pressure data uncertainty in WDNs. An artificial neural network (ANN) is used to predict nodal pressures and measurement errors are generated by using cumulative density function inverse sampling method that follows Gaussian distribution. Total of nine conditions (3 input datasets × 3 output datasets) are considered in the ANN model to investigate the impact of measurement error size on the prediction results. The results have shown that higher data uncertainty decreased ANN model's prediction accuracy. Also, the measurement error of output data had more impact on the model performance than input data that for a same measurement error size on the input and output data, the prediction accuracy was 72.25% and 38.61%, respectively. Thus, to increase ANN models prediction performance, reducing the magnitude of measurement errors of the output pressure node is considered to be more important than input node.

안정적인 수도 공급을 위한 상수도관망의 역할이 더욱 주목받음에 따라 비정상 상황에 대한 신속한 탐지와 적절한 대처 역시 중요시되고 있다. 장치에 의존한 탐지기법 등 기존의 방법론에는 한계가 존재하므로 데이터를 이용한 모델 기반의 방법이 개발되었다. 하지만 상수도관망 내 측정 데이터는 불확실성을 가져 실제 사용량과 다르다. 따라서 본 연구에서는 기계학습 방법의 하나인 인공신경망 모델을 이용하여 상수도관망 압력값을 예측함에 있어 데이터 불확실성의 영향을 조사한다. 정규분포를 따르는 임의의 값을 고려하여 데이터에 측정치 오류를 형성하고 측정치 오류 여부 및 종류에 따라 총 9가지 데이터를 인공신경망 모델을 통해 예측해 경향성을 비교한다. 분석을 통해 데이터 불확실성이 증가할수록 모델 성능이 감소하며, 출력데이터의 측정치 오류가 모델 성능에 미치는 정도가 더 큼을 확인하였다. 특히 입력데이터와 출력데이터의 측정 오차 크기가 동일한 경우 예측 정확도는 각각 72.25%, 38.61%로 큰 차이를 보였다. 따라서 ANN 모델 예측 성능 향상을 위해서는 입력 데이터보다 출력데이터인 주절점의 측정 오류 크기를 줄이는 것이 중요하다.

Keywords

Acknowledgement

이 성과는 정부(과학기술정보통신부)의 재원으로 한국연구재단의 지원을 받아 수행된 연구임(No. NRF-2021R1A5A1032433).

References

  1. Badger Meter, Inc. (2022). E-series ultrasonic meters for residential applications, accessed 19 October 2022, .
  2. Gurung, T.R., Stewart, R.A., Beal, C.D., and Sharma, A.K. (2015). "Smart meter enabled water end-use demand data: Platform for the enhanced infrastructure planning of contemporary urban water supply networks." Journal of Cleaner Production, Vol. 87, pp. 642-654. https://doi.org/10.1016/j.jclepro.2014.09.054
  3. Jun, S., and Choi, Y.H. (2022). "Data generation approaches to detect abnormal conditions in water distribution systems. 2." Journal of Korea Society of Hazard Mitigation, Vol. 22, pp. 69-79.
  4. Jun, S., Jung, D., and Lansey, K.E. (2021). "Comparison of imputation methods for end-user demands in water distribution systems." Journal of Water Resources Planning and Management, Vol. 147, No. 12, 04021080.
  5. Jung, D., and An, J. (2018). "Data-driven event detection method for efficient management and recovery of water distribution system man-made disasters." Journal of Korea Water Resources Association, Vol. 51, No. 8, pp. 703-711.
  6. Jung, D., and Lansey, K. (2014). "Burst detection in water system using the extended Kalman filter." Procedia Engineering, Vol. 70, pp. 902-906. https://doi.org/10.1016/j.proeng.2014.02.100
  7. Jung, D., Kang, D., Liu, J., and Lansey, K. (2015). "Improving the rapidity of responses to pipe burst in water distribution systems: A comparison of statistical process control methods." Journal of Hydroinformatics, Vol. 17, No. 2, pp. 307-328. https://doi.org/10.2166/hydro.2014.101
  8. Kim, S., Choi, D., Bae, C., and Kim, J. (2013). "Leakage detection of water distribution system using adaptive Kalman filter." Journal of Korea Water Resources Association, Vol. 46, No. 10, pp. 969-976. https://doi.org/10.3741/JKWRA.2013.46.10.969
  9. Lee, C.W., and Yoo, D.G. (2021). "Development of leakage detection model in water distribution networks applying LSTM-based deep learning algorithm." Journal of Korea Water Resources Association, Vol. 54, No. 8, pp. 599-606.
  10. Lee, E., and Coi, B. (2018). "Transformation of paradigm of resource security through water-energy-food nexus." Journal of the Korean Society of Civil Engineers, Vol. 66, No. 6, pp. 16-20.
  11. Mounce, S.R., and Machell, J. (2006). "Burst detection using hydraulic data from water distribution systems with artificial neural networks." Urban Water Journal, Vol. 3, No. 1, pp. 21-31. https://doi.org/10.1080/15730620600578538
  12. Ozger, S.S. (2003). A semi-pressure-driven approach to reliability assessment of water distribution networks. Ph. D. dissertation, University of Arizona, AZ, U.S.
  13. Park, G.Y., Choe, J.H., and Jeong, D.H. (2020). "A study on the practical use of smart meter end-user demand data." Water for Future, Vol. 53, No. 5, pp. 77-81.
  14. Rossman, L.A. (2000). EPANET 2 user's manual. United States Environmental Protection Agency, Cincinnati, OH, U.S.
  15. Sanneh, J., Di Mauro, A., Venticinque, S., Santonastaso, G.F., Di Nardo, A., Wang, Y., and Lee, J. (2022). "Water demand forecasting time series data." Embracing analytics in the drinking water industry, Editied by Lee, J., Keck, J., IWA Publishing, London, pp.75-122.
  16. Wan, X., Kuhanestani, P.K., Farmani, R., and Keedwell, E. (2022). "Literature review of data analytics for leak detection in water distribution networks: A focus on pressure and flow smart sensors." Journal of Water Resources Planning and Management, Vol. 148, No. 10, 03122002.
  17. Ye, G., and Fenner, R.A. (2011). "Kalman filtering of hydraulic measurements for burst detection in water distribution systems." Journal of Pipeline Systems Engineering and Practice, Vol. 2, No. 1, pp. 14-22. https://doi.org/10.1061/(ASCE)PS.1949-1204.0000070