تحلیل خشکسالی هواشناسی با استفاده از الگوریتم بهینه‌سازی ازدحام ذرات- شبکه‌های عصبی مصنوعی بر مبنای شاخص MSPI

نویسندگان

1 دانش‌آموخته کارشناسی ارشد، گروه مهندسی عمران، پردیس خودگردان تبریز، دانشگاه تبریز

2 استاد، گروه مهندسی عمران آب، دانشکده مهندسی عمران، دانشگاه تبریز

3 دانش‌آموخته دکتری، گروه مهندسی عمران آب، دانشکده مهندسی عمران، دانشگاه تبریز

چکیده

پدیده خشکسالی یکی از بلایای طبیعی می­باشد که احتمال وقوع آن در تمام مناطق اقلیمی امکان­پذیر است و در هر منطقه­ای که روی می­دهد، باعث ایجاد آسیب­های جدی در محیط زیست و زندگی انسان­ها می­شود. بنابراین، پیش­بینی این پدیده مضر، می­تواند تأثیر قابل توجهی در مدیریت منابع آب داشته باشد و آثار مخرب آن را تا حد امکان کاهش دهد. در این مطالعه، ابتدا با استفاده از شاخص بارندگی استانداردشده چند متغیره (MSPI)، مشخصه­های خشکسالی در حوضه آبریز لیقوان­چای به­دست آمد و سپس از شبکه­های عصبی مصنوعی (ANN) جهت پیش­بینی شاخص فوق استفاده گردید. جهت آموزش شبکه­های عصبی مصنوعی و تخمین بهینه وزن­های آن، الگوریتم بهینه­سازی ازدحام ذرات (PSO) به­کار برده شد و عملکرد آن با الگوریتم پس انتشار خطا (BP) مورد مقایسه قرار گرفت. در این راستا سناریوها و ساختارهای مختلفی در نظر گرفته شد و سپس با استفاده از آزمون­های نیکوئی برازش، میزان دقت هر یک از آن­ها محاسبه گردید.  نتایج حاصل، برتری مدل ANN-PSO نسبت به مدل ANN-BP در پیش­بینی خشکسالی را نشان داد.

کلیدواژه‌ها


عنوان مقاله [English]

Meteorological Drought Analysis Using Particle Swarm Optimization Algorithm- Artificial Neural Networks Based on MSPI Index

نویسندگان [English]

  • M Shafiei Najd 1
  • Y Hassanzadeh 2
  • MT Alami 2
  • A Abdi Kordani 3
1 Graduate, Civil Engin. Dept., Tabriz Campus, Univ. of Tabriz, Iran
2 Prof., Water Engin. Dept., Faculity of Civil Engin., Univ. of Tabriz, Iran
3 Graduate, Water Engin. Dept., Faculity of Civil Engin., Univ. of Tabriz, Iran
چکیده [English]

The drought phenomenon is one of the natural disasters, which may occur in all climatic zones and cause serious damages to the environment and human life. So, forecasting this phenomenon may have significant impact on the water resources management and reduce its destructive effects as much as possible. In this study, the multivariate standardized precipitation index (MSPI) was utilized to compute the drought characteristics in the Lighvanchai basin and then the artificial neural network (ANN) was used to forecast the MSPI values. In order to train the ANN and estimate its optimized weights, the particle swarm optimization (PSO) algorithm was applied and its performance was compared with the backpropagation (BP) algorithm. In this context, different scenarios and structures were considered and then the goodness-of-fit tests were utilized for evaluating the accuracy of them. The results demonstrated that the ANN-PSO model had a better performance than the ANN-BP model for drought forecasting.

کلیدواژه‌ها [English]

  • Artificial neural network
  • Lighvan Chai Basin
  • Meteorological drought
  • Multivariate standardized precipitation index
  • Particle swarm optimization algorithm
Asadnia M, Lioyd HC, Chau XS, Qin AM and Talei A, 2014. Improved particle swarm optimization-based artificial neural network for rainfall-runoff modeling. Journal of Hydrologic Engineering 19: 1320-1329.
Bacanli UG, Firat M and Dikbas F, 2009. Adaptive neuro-fuzzy inference system for drought forecasting. Stochastic Environmental Research and Risk Assessment 23: 1143-1154.
Bazrafshan J, Hejabi S and Rahimi J, 2014. Drought monitoring using the multivariate standardized precipitation index (MSPI). Water Resources Management 28: 1045-1060.
Bazrafshan J, Nadi M and Ghorbani K, 2015. Comparison of empirical copula-based joint deficit index (JDI) and multivariate standardized precipitation index (MSPI) for drought monitoring in Iran. Water Resources Management 29(6): 2027-2044.
Belayneh A, Adamowski J, Khalil B and Ozga-Zielinski B, 2014. Long-term SPI drought forecasting in the Awash River Basin in Ethiopia using wavelet neural network and wavelet support vector regression models. Journal of Hydrology 508: 418-429.
Chau KW, 2006. Particle swarm optimization training algorithm for ANNs in stage prediction of Shing Mun River. Journal of Hydrology 329: 363-367.
Eberhart RC and Shi Y, 2000. Comparing inertia weights and constriction factors in particle swarm optimization. Pp. 84-88. Proceedings of the Congress on Evolutionary Computation, IEEE, La Jolla, California, USA.
Hassanzadeh Y, Abdi A and Fakheri-Fard A, 2012. Drought forecasting using genetic algorithm and conjoined model of neural network-wavelet. Water and Wastewater 23(83): 48-59.
Hassanzadeh Y, Abdi A, Shafiei M and Khoshtinat S, 2015. Daily streamflow forecasting of Nooranchay River using the hybrid model of artificial neural networks-principal component analysis. Water and Soil Science- University of Tabriz 25(3): 53-63.
Hassanzadeh Y, Abdi A, Talatahari S and Singh VP, 2011. Meta-heuristic algorithms for hydrologic frequency analysis. Water Resources Management 25(7): 1855-1879.
Kennedy J and Eberhart RC, 1995. Particle swarm optimization. Pp. 1942-1948. Proceedings of the 1995 IEEE International Conference on Neural Networks, IEEE Service Center, Perth, Australia.
Kim TW and Valdes JB, 2003. Nonlinear model for drought forecasting based on a conjunction of wavelet transforms and neural networks. Journal of Hydrologic Engineering 8(6): 319-328.
Mirjalili S, Mohd Hashim SZ and Sardroudi HM, 2012. Training feedforward neural networks using hybrid particle swarm optimization and gravitational search algorithm. Applied Mathematics and Computation 218: 11125-11137.
Mishra AK and Desai VR, 2006. Drought forecasting using feed-forward recursive neural network. Ecological Modelling 198: 127-138.
Morid S, Smakhtin V and Bagherzadeh K, 2007. Drought forecasting using artificial neural networks and time series of drought indices. International Journal of Climatology 27: 2103-2111.
Piotrowski AP and Napiorkowski JJ, 2011. Optimizing neural networks for river flow Forecasting-Evolutionary Computation methods versus the Levenberg-Marquardt approach. Journal of Hydrology 407: 12-27.
Rezaeian-Zadeh M and Tabari H, 2012. MLP-based drought forecasting in different climatic region. Theoretical Applied Climatology 109: 407-414.
Sharma S, 1996. Applied multivariate techniques. John Wiley & Sons, New York.
Shi Y and Eberhart R 1998. A modified particle swarm optimizer. Pp. 69-73. Proceedings of the IEEE International Conference on Evolutionary Computation, IEEE Computer Society, Washington, USA.
Topoglou E, Trichakis IC, Doku Z, Nikolos IK and Karatsaz GP, 2014. Groundwater-level forecasting under climate change scenarios using an artificial neural network trained with particle swarm optimization. Hydrological Sciences Journal 59(6): 1225-1239.
Wilks DS, 2011. Statistical methods in the atmospheric sciences, third edition, Academic Press, Amsterdam.
Zhang JR, Zhang J, Lok TM and Lyu MR, 2007. A hybrid particle swarm optimization–back-propagation algorithm for feedforward neural network training. Applied Mathematics and Computation 185: 1026-1037.