ارزیابی ترکیب الگوریتم‌های بهینه‌سازی و سیستم استنتاج فازی- عصبی تطبیقی در مقایسه با مدل‌های سری‌های زمانی در تخمین سطح آب زیرزمینی

نوع مقاله : مقاله پژوهشی

نویسندگان

1 دانشگاه تهران

2 دانشجوی دکتری تخصصی مهندسی منابع آب،گروه مهندسی آبیاری و آبادانی دانشگاه تهران،کرج،ایران

3 دانشکده جغرافیا، دانشگاه تهران

چکیده

 
به‌منظور مدیریت و بهره­برداری بهینه از منابع آب زیر‌زمینی آگاهی از تغییرات مکانی- زمانی سطح ایستابی و لزوم پیش­بینی و مدل‌سازی آنها به‎منظور شناخت دقیق­تر رفتار آبخوان نسبت به محرک­های طبیعی و انسانی، امری ضروری است. با توجه به توسعه روزافزون فرامدل­ها و ترکیب آنها با الگوریتم­های بهینه­سازی به منظور مدل‌سازی و پیش­بینی متغیرهای هیدروژئولوژیکی، این سؤال که استفاده از مدل­های ترکیبی چقدر می‌تواند عملکرد فرامدل‌ها را بهبود بخشد، مطرح می‌شود. به منظور تلاشی در جهت یافتن پاسخ، در این پژوهش، چهار الگوریتم بهینه‌سازی فراکاوشی ازدحام ذرات (PSO)، ژنتیک (GA)، کلونی مورچگان (ACOR) و تکاملی تقاضلی (DE) با مدل سیستم استنتاج فازی- عصبی تطبیقی (ANFIS) ترکیب شد. عملکرد چهار مدل­ ترکیبی توسعه داده شده با مدل ANFIS و مدل سری زمانی (SARIMA) به عنوان مدل­ مرجع، جهت تخمین سطح آب زیرزمینی متوسط ماهانه آبخوان دشت صحنه در استان کرمانشاه، در بازه زمانی 19 سال آبی ارزیابی شد. به‎منظور مقایسه بهتر نتایج مدل­ها، متغیرهای ورودی یکسان از تراز آب زیرزمینی در گام­های زمانی مختلف (حداکثر 4 ماه بر اساس تابع خودهمبستگی تراز آبخوان) برای آنها درنظر گرفته شد. نتایج شاخص­های نکویی برازش در مرحله آموزش و آزمون نشان داد اختلاف معنا‌داری بین مدل سری زمانی SARIMA نسبت به سایر مدل‌های ترکیبی مورد استفاده، وجود ندارد. اما با توجه به اینکه SARIMA فرآیندهای میانگین متحرک، اتورگرسیون، تغییرات فصلی و تأخیر را در مدل‌سازی اعمال می‌کند، در مدل‌سازی سطح آب زیرزمینی می­تواند بیشتر مورد توجه قرار گیرد. مقادیر RMSE برترین مدل‌ ترکیبی (ANFIS-GA) و SARIMA به ترتیب 0950/0 و 1012/0 متر به دست آمد. همچنین نتایج به دست آمده نشان داد که ترکیب الگوریتم­های بهینه‌سازی درنظر گرفته شده با مدل ANFIS نتایج مدل را نسبت به مدل انفرادی ANFIS به­صورت معنی‌داری بهبود نمی­بخشد. نتایج این تحقیق می­تواند محققان را در انتخاب آگاهانه مدل مناسب در پیش­بینی زمانی سطح ایستابی آبخوان با توجه به معیارهای کارآیی، زمان و هزینه محاسبات و آماده­سازی داده­ها جهت ورود به مدل­ها کمک نماید.

کلیدواژه‌ها

موضوعات


عنوان مقاله [English]

Evaluation of the combination of optimization algorithms and adaptive fuzzy-neural inference system compared to time series models in groundwater level estimation

نویسندگان [English]

  • masoumeh zeinalie 1
  • mohammad ansari ghojghar 2
  • yaser mehri 1
  • seiyed mossa hosseini 3
1 tehran university
2 Phd Candidate of Water Resources Engineering, Department of Irrigation and Reclamation Engineering, University of Tehran, Karaj, Iran.
3 tehran university
چکیده [English]

To optimize the management and optimal use of groundwater resources, it is necessary to be aware of the temporal-spatial changes of the stagnant level . For modeling and predicting hydrogeological variables, the question remains:To what extent these hybrid models can be effective compared to the individual model?, in this study four algorithms of particle overvoltage optimization (PSO) genetics (GA) ant colony (ACOR) and demand evolution (DE) were combined with the model of adaptive fuzzy-neural inference system (ANFIS).The four combined models performance developed with the ANFIS model and the time series model (SARIMA) as the reference model to estimate the average monthly groundwater level of the Sahneh plain aquifer in Kermanshah province was evaluated over 19 years.To better compare the results of the models, the same input variables of the groundwater level in different time steps (maximum four months based on the self-correlation function of aquifer level) were considered for them. The results of fitness indicators in the test and test phase showed that there was no significant difference between the SARIMA time series model compared to other combined models used.However, given that SARIMA applies average moving processes, authorization, seasonal changes, and delays in modeling, groundwater leveling can be given more attention in modeling. The RMSE values of the best hybrid model (ANFIS-GA) and SARIMA were 0.950 and 0.1012, respectively. The results also showed that the combination of optimization algorithms considered with the ANFIS model does not improve the model's results compared to the individual ANFIS model in terms of significance.

کلیدواژه‌ها [English]

  • Modeling
  • groundwater level
  • SARIMA
  • ANFIS
  • evolutionary algorithms
Araghinejad S (2013) Data-driven modeling: Using MATLAB® in water resources and environmental engineering (Vol. 67). Springer Science & Business Media
Barzegar R, Fijani E, Asghari Moghaddama A, Tziritis E (2017) Forecasting of groundwater level fluctuations using ensemble hybrid multi-wavelet neural network-based models. Science of the Total Environment 599(60):20–31
Dashti R, Sattari M T, and Nourani V (2017) Performance evaluation of differential evolution algorithm in optimum operating of Eleviyan single-reservoir dam system. Journal of Water and Soil Resources Conservation 6(3):61-75
Deb K A P, Agarwal S, Meyarivan T (2000) A fast elitist multi-objective genetic algorithm: NSGA II. IEEE Transactions on Evolutionary Computation 6:182-197
Dorigo M (1992) Optimization, learning and natural algorithms. Ph.D. Thesis, Dipartimento di Elettronica, Politecnico di Milano, Italy
Ghadmapour Z and Shaghaghian M (2011) Comparison of classical models of time series and artificial intelligence in determining groundwater levels. 6th National Congress of Civil Engineering, Semnan, Semnan University, https://www.civilica.com/Paper-NCCE06- NCCE06_0809.html
Habibi M H, Nadiri A, and Asghari Moghadam A (2016) Providing a combined genetic-kriging model for temporal and spatial prediction of groundwater level. Iran-Water Resources Research 11(3):85-99 (In Persian)
Hosseini M and Kerachian R (2017) A data fusion-based methodology for optimal redesign of groundwater monitoring networks. Journal of Hydrology 552:267-282
Hosseini S M & Mahjouri N (2014) Developing a fuzzy neural network-based support vector regression (FNN-SVR) for regionalizing nitrate concentration in groundwater. Environmental Monitoring and Assessment 186(6):3685-3699
Hosseinipoor A, and Abdollahi A (1991) Predicting the short-term load of Kerman region using ANFIS-PSO combined method. Second National Conference on New and Clean Energy Management, Shahid Ba Honar University of Kerman
Jalalkamali A (2015) Using of hybrid fuzzy models to predict spatiotemporal groundwater quality parameters. Earth Science Informatics 8(4):885-894
Jang J S R, Sun C T, and Mizutani E (1997) Neuro-fuzzy and soft computing: A computational approach to learning and machine intelligence. Prentice-Hall International, New Jersey
Kholghi M and Hosseini S M (2009) Comparison of groundwater level estimation using neuro-fuzzy and ordinary kriging. Environmental Modeling and Assessment 14(6):729
Kişi Ö (2009) Evolutionary fuzzy models for river suspended sediment concentration estimation. Journal of Hydrology 372(1–4):68-79
Ljung G M and Box G E P (1978) On a measure of lack affect in time series models. Biometrica 65(2):297-303
Malekinezhad H and Pourshareyati R (2013) Application and comparison of cumulative time series model and artificial neural network model in prediction of groundwater level variation (Case study: Marvast plain). Irrigation Science & Engineering (Scientific Journal of Agriculture) 36(3):81-92 (In Persian)
Nadiri A, Taherkhani Z, and Sadeghi Aghdam F (2017) Groundwater level prediction of Bostanabad plain using supervised combination of artificial intelligence models. Iranian Water Resources Research 13(3):43-55 (In Persian)
Naseri A (2018) Comparison of the application of fourteen temporal series patterns to analyze and predict changes in groundwater level in Marand plain (northern margin of Urmia Lake). Irrigation and Drainag 13(1):58-68
Nayak P C, Sudheer K P, Rangan D M, and Ramasastri K S (2004) A neuro-fuzzy computing technique for modeling hydrological time series. Journal of Hydrology 291(1–2):52-66
Nourani V, Salehi K (2008) Rainfall-runoff modeling using adaptive fuzzy neural network method and comparing it with neural network and fuzzy inference method, Case study: (Lighvan Chay catchment area in East Aegean province). Fourth National Congress on Civil Engineering, University of Tehran (In Persian)
Pour Mohammadi S, Malekinejad A, Pourshareiani R (2013) Comparison of the efficiency of neural and serial time network method in predicting groundwater level Case study: Under Bakhtegan basin of Fars province. Journal of Water and Soil Research 20(4):251:262 (In Persian)
Radmanesh F, Golabi M R, Khodabakhshi F, Farzi S, and Zeinali M (2020) Modeling aquifer hydrograph: Performance review of conceptual MODFLOW and simulator models. Arabian Journal of Geosciences 13(5):1-9
Safavid H (2011) Engineering hydrology. Arkan Danesh Publications, Third Edition, 724 pages
Shirvanyichi H, Rashkolia H, Zonamat Kermani M (2013) Simulation and prediction of the groundwater status of Kerman plain using return and time series models of Kerman stochastic using return models and time series of stochastic time. The First National Conference on Hydrology in Arid Regions, 3 to 5 May, Kurdistan University Jihad
Socha K, Dorigo M (2008) Ant colony optimization for continuous domains. European Journal of Operational Research 185(3):1155-1173
Stanley Raj A, Hudson Oliver D,  Srinivas Y, Viswanath J (2017) Wavelet based analysis on rainfall and water table depth forecasting using Neural Networks in Kanyakumari district, Tamil Nadu, India. Groundwater for Sustainable Development 5:178–186
Storn R, Price K (1997) Differential evolution-A simple and efficientheuristic for global optimization over continuous spaces. Journal of Global Optimization 11:341-359
Wang Sh, Feng J, Liu G (2013) Application of seasonal time series model in the precipitation forecast. Mathematical and Computer Modelling 58(3-4):677-683
Yoona H, Hyunb Y, Ha K, Leec K K, Kimd G B ( 2016) A method to improve the stability and accuracy of ANN- and SVM-based time series models for long-term groundwater level predictions. Computers and Geosciences 90:144-155
Zahraei B, Hosseini M (2014) Genetic algorithm and engineering optimization. Publications: Gutenberg, Second Edition, 298 pages
Zareie H, Akhondali A M (2007) Comparison of ground-statistical methods and classical statistics in hydrograph drawings of groundwater unit. Research Project of Khuzestan Water and Power Organization P. 71 (In Persian)