بررسی اثر استفاده از تبدیلات موجک بر روی عدم‌قطعیت مدل‌های مبتنی بر شبکه عصبی مصنوعی و ماشین یادگیری افراطی در زمینه پیش‌بینی میزان تقاضای آب شرب شهری

نوع مقاله : مقاله پژوهشی

نویسندگان

1 دانش‌آموخته کارشناسی ارشد مهندسی عمران/ محیط زیست، گروه مهندسی عمران، دانشکده فنی و مهندسی، دانشگاه صنعتی قم، قم، ایران.

2 استادیار/ گروه مهندسی عمران، دانشکده فنی و مهندسی، دانشگاه صنعتی قم، قم، ایران.

3 استادیار/ گروه مهندسی کامپیوتر، دانشکده فنی و مهندسی، دانشگاه صنعتی قم، قم، ایران.

چکیده

پیش‌بینی میزان مصرف آب شرب شهری یکی از دغدغه‌های نوین جوامع شهری معاصر بوده است. در این راستا، تحقیقات زیادی در زمینه مقایسه عملکرد مدل‌های مختلف انجام شده است. با معرفی شبکه عصبی مصنوعی، بحث پیرامون نحوه بهینه‌سازی آن‌ها با استفاده از روش‌های مختلف، بخصوص تبدیلات موجک داغ شد. در اغلب پژوهش‌ها اثر استفاده از تبدیلات موجک بر روی عملکرد و دقت مدل‌های عصبی مورد توجه قرار گرفت، اما تاثیر استفاده از تبدیلات موجک بر عدم‌قطعیت مدل‌های عصبی مورد بررسی قرار نگرفته است. در این پژوهش عملکرد و عدم قطعیت دو مدل مبتنی‌بر شبکه عصبی مصنوعی بازگشتی (NARX)، مدل یادگیری ماشینی افراطی (ELM) و نسخه موجکی آن‌ها (W_NARX) و (W_ELM) برای پیش‌بینی میزان مصرف آب شهرک مهدیه قم مورد بررسی قرار گرفت. نتایج نشان داد که مدل NARX (با ضریب رگرسیون ۰.۹۵۵) از دقت بالاتری در مقایسه با ELM (با ضریب رگرسیون ۰.۷۸۷) برخوردار است. از طرفی، نوع موجکی آن‌ها به‌ترتیب دارای ضریب رگرسیون ۰.۹۶۰ و ۰.۸۴۷ است که نشان دهنده برتری مدل W_NARX است. علت عملکرد ضعیف‌تر ELM را می‌توان در پیچیدگی زیاد رفتار مصرف‌کننده آب و ساختار ساده این مدل نسبت به NARX دانست. از طرفی، استفاده از تبدیلات موجک بر بهبود دقت هر دو مدل تاثیر مثبت داشت، اما این تاثیر در مدل ELM بیشتر بود. نتایج تحلیل عدم‌قطعیت بر روی این دو مدل حاکی از کاهش عدم‌قطعیت هر دو مدل بود. اما این مهم در مدل W_NARX با بازه اطمینان ۹۸.۷۵٪ بیشتر قابل توجه بود.

کلیدواژه‌ها


عنوان مقاله [English]

Studying the Effect of Wavelet Transform on the Uncertainty of Artificial Neural Network-based Models and Extreme Learning Machines for the Prediction of Urban Water Demand

نویسندگان [English]

  • Mostafa Rezaali 1
  • Abdolreza Karimi 2
  • Bayraali Mohammadnezhad 2
  • Abdolreza Rasouli 3
1 M.Sc. Graduate of Civil and Environmental Engineering, Department of Civil Engineering, Qom University of Technology (QUT), Qom, Iran
2 Assistant Professor, Department of Civil Engineering, Qom University of Technology (QUT), Qom, Iran.
3 Assistant Professor, Department of Computer Engineering, Qom University of Technology (QUT), Qom, Iran.
چکیده [English]

Urban water demand prediction has been one of the contemporary concerns of modern urban societies. In this vein, many studies have been carried out comparing the performance of different models. By the introduction of artificial neural networks (ANNs), the discussion about the accuracy improvement of ANNs using wavelet transforms (WTs) was heated up. In many research, the effect of using WTs on the performance and the accuracy of ANNs drew many attentions. However, the effect of using WTs on the uncertainties associated with ANNs has not been investigated. In this study, the performance and the uncertainty of two ANN-based models, i.e., nonlinear autoregressive network with exogenous inputs (NARX) and extreme learning machines (ELM) were studied and the wavelet version of those, i.e., W_NARX and W_ELM were used for the prediction of urban water demand of Mahdie Residential Complex. The results indicated that NARX (regression coefficient (R) of 0.955) is more accurate than ELM (R of 0.787). On the other, the WT version of these models had the R of 0.960 and 0.847, respectively, indicating the outperformance of W_NARX model. The reason for the lower accuracy of ELM could be found in the complexity of the water consumer behavior and the simpler structure of ELM than NARX. Besides, the implementation of WTs had a positive effect on both models, but ELM more. The results of the uncertainty analysis of both models indicated a decrease in uncertainty. However, this was more considerable in W_NARX with the confidence interval of 98.75%.

کلیدواژه‌ها [English]

  • Artificial Neural Networks
  • Extreme Learning Machines
  • wavelet transform
  • Uncertainty
Adamowski J, Fung Chan H, Prasher SO, Ozga‐Zielinski B and Sliusarieva A (2012) Comparison of multiple linear and nonlinear regression, autoregressive integrated moving average, artificial neural network, and wavelet artificial neural network methods for urban water demand forecasting in Montreal, Canada. Water Resources Research 48(1)
Adamowski J and Karapataki C (2010) Comparison of multivariate regression and artificial neural networks for peak urban water-demand forecasting: evaluation of different ANN learning algorithms. Journal of Hydrologic Engineering 15(10):729–743
Adamowski JF (2008) Peak daily water demand forecast modeling using artificial neural networks. Journal of Water Resources Planning and Management 134(2):119–128
Alexandridis AK and Zapranis AD (2013) Wavelet neural networks: A practical guide. Neural Networks 42:1–27
Antonini M, Barlaud M, Mathieu P and Daubechies I (1992) Image coding using wavelet transform. IEEE Transactions on Image Processing 1(2):205–220
Babel MS and Shinde VR (2011) Identifying prominent explanatory variables for water demand prediction using artificial neural networks: a case study of Bangkok. Water Resources Management 25(6):1653–1676
Basheer IA and Hajmeer M (2000) Artificial neural networks: fundamentals, computing, design, and application. Journal of Microbiological Methods 43(1):3–31
Bašta M (2014) Additive decomposition and boundary conditions in wavelet-based forecasting approaches. Acta Oeconomica Pragensia 22(2):48–70
Campisi-Pinto S, Adamowski J and Oron G (2012) Forecasting urban water demand via wavelet-denoising and neural network models. Case study: city of Syracuse, Italy. Water Resources Management 26(12):3539–3558
Daubechies I (1990) The wavelet transform, time-frequency localization and signal analysis. IEEE Transactions on Information Theory 36(5):961–1005
Duerr I, Merrill HR, Wang C, Bai R, Boyer M, Dukes MD and Bliznyuk N (2018) Forecasting urban household water demand with statistical and machine learning methods using large space-time data: A Comparative study. Environmental Modelling & Software 102:29–38
Galelli S and Castelletti A (2013) Tree-based iterative input variable selection for hydrological modeling. Water Resources Research 49(7):4295–4310
Gençay R and Liu T (1997) Nonlinear modelling and prediction with feedforward and recurrent networks. Physica D: Nonlinear Phenomena 108(1–2):119–134
Ghiassi M, Zimbra DK, and Saidane H (2008) Urban water demand forecasting with a dynamic artificial neural network model. Journal of Water Resources Planning and Management 134(2):138–146
Gupta H V, Hsu K, and Sorooshian S (2000) Effective and efficient modeling for streamflow forecasting. Artificial neural networks in hydrology. Springer, 7–22
Herrera M, Torgo L, Izquierdo J, and Pérez-García R (2010) Predictive models for forecasting hourly urban water demand. Journal of Hydrology 387(1–2):141–150
Huang G-B, Zhu Q-Y and Siew C-K (2006) Extreme learning machine: theory and applications. Neurocomputing 70(1–3):489–501
Joo CN, Koo JY and Yu MJ (2002) Application of short-term water demand prediction model to Seoul. Water Science and Technology 46(6–7):255–261
Lin T, Horne BG, Tino P, and Giles CL (1996) Learning long-term dependencies in NARX recurrent neural networks. IEEE Transactions on Neural Networks 7(6):1329–1338
Maier HR and Dandy GC (2000) Neural networks for the prediction and forecasting of water resources variables: a review of modelling issues and applications. Environmental Modelling & Software 15(1):101–124
Maslova I, Ticlavilca AM, and McKee M (2016) Adjusting wavelet‐based multiresolution analysis boundary conditions for long‐term streamflow forecasting. Hydrological Processes 30(1):57–74
Percival DB (2008) Analysis of geophysical time series using discrete wavelet transforms: an overview. Nonlinear Time Series Analysis in the Geosciences. Springer, 61–79
Pittner S and Kamarthi S V (1999) Feature extraction from wavelet coefficients for pattern recognition tasks. IEEE Transactions on Pattern Analysis urban water consumption and weather variables in Seoul, Korea. Physical Geography 30(4) and Machine Intelligence 21(1):83–88
Praskievicz S and Chang H (2009) Identifying the relationships between:324-337
Quilty J and Adamowski J (2018) Addressing the incorrect usage of wavelet-based hydrological and water resources forecasting models for real-world applications with best practices and a new forecasting framework. Journal of Hydrology 563:336–353
Refenes AN, Zapranis A, and Francis G (1994) Stock performance modeling using neural networks: a comparative study with regression models. Neural Networks 7(2):375–388
Rezaali M, Karimi A, Mohammadnezhad B, and Rasouli A (2018) Comparison of methods for urban water demand prediction, case study: Mahdie Residential Complex, Qom. Second Conference of Water and Wastewater Engineering. Isfahan University of Technology: Iran Water and Wastewater Organization. Available at: https://www.civilica.com/Paper-NWWCE02-NWWCE02_014.html (In Persian)
Santoso S, Powers EJ, and Grady WM (1997) Power quality disturbance data compression using wavelet transform methods. IEEE Transactions on Power Delivery 12(3):1250–1257
Schalkoff R (1992) Pattern recognition: Statistical, structural and neural approaches. John Wiley & Sons. Inc, New York
Sifuzzaman M, Islam MR, and Ali MZ (2009) Application of wavelet transform and its advantages compared to Fourier transform. Journal of Physical Sciences 13:121-134
Tajbakhsh A, Nourani V, Molajou A (2019) Hybrid Wavelet-M5 modeling in rainfall-runoff process forecast. Iran-Water Resources Research 15(2):1735-2347 (In Persian)
Vemuri R V and Rogers RD (1994) Artificial neural networks: Forecasting time series. Los Alamitos, Calif. : IEEE Computer Society Press, c1994
Wan C, Xu Z, Pinson P, Dong ZY, and Wong KP (2014) Probabilistic forecasting of wind power generation using extreme learning machine. IEEE Transactions on Power Systems 29(3):1033-1044
Zhou SL, McMahon TA, Walton A, and Lewis J (2000) Forecasting daily urban water demand: a case study of Melbourne. Journal of Hydrology 236(3–4):153–164