تخمین خشکسالی دراستان لرستان با استفاده از شبکه های هوشمند

نوع مقاله : مقاله پژوهشی

نویسندگان

1 دانشیار، گروه مهندسی آب ، دانشگاه لرستان

2 دانشجوی دکترای سازه های آبی، دانشگاه لرستان

3 دانشجوی دکترای سازه آبی

چکیده

خشکسالی یکی از پدیده‌های آب و هوایی است که در همه شرایط اقلیمی و در همه مناطق کره زمین به وقوع می‌پیوندد. پیش‌بینی خشک‌سالی نقش مهمی در طراحی و مدیریت منابع طبیعی، سیستم‌های منابع آب، تعیین نیاز آبی گیاه  ایفا می‌نماید. در این پژوهش جهت تخمین شاخص بارش استاندارد 12 ماهه چهار ایستگاه باران سنجی نورآباد، الشتر، درود و بروجرد واقع در استان لرستان از مدل شبکه عصبی موجک استفاده شد و نتایج آن با سایرروشهای هوشمند از جمله  شبکه عصبی مصنوعی مقایسه گردید. برای این منظور از پارامتر بارش در مقیاس زمانی ماهانه در طی دوره آماری (1372-1392) بعنوان ورودی و شاخص بارش استاندارد بعنوان پارامتر خروجی مدلها انتخاب گردید. معیارهای ضریب همبستگی، ریشه میانگین مربعات خطا و میانگین قدر مطلق خطا برای ارزیابی و عملکرد مدلها مورد استفاده قرار گرفت. نتایج نشان داد هر دو مدل قابلیت خوبی در تخمین شاخص بارش استاندارد دارند، لیکن از لحاظ دقت، مدل شبکه عصبی موجک در ایستگاه دورود ضریب همبستگی 811/0 و کمترین ریشه میانگین مربعات خطا 068/0 میلی متر و کمترین میانگین قدر مطلق خطا 051/0 میلی متر، در ایستگاه بروجرد ضریب همبستگی 885/0 و کمترین ریشه میانگین مربعات خطا 056/0 میلی متر و کمترین میانگین قدر مطلق خطا 048/0 میلی متر، ایستگاه الشتر  ضریب همبستگی 827/0 و کمترین ریشه میانگین مربعات خطا 045/0 میلی متر و کمترین میانگین قدر مطلق خطا 039/0 میلی متر و در نهایت در ایستگاه نورآباد با ضریب همبستگی 849/0 و کمترین ریشه میانگین مربعات خطا 050/0 میلی متر و کمترین میانگین قدر مطلق خطا 046/0 میلی متر در مرحله صحت سنجی نسبت به سایر ساختارها جهت مدل‌سازی شاخص بارش استاندارد درمقیاس زمانی ماهانه عملکرد بهتری نسبت به شبکه عصبی مصنوعی از خود نشان داده است. در مجموع نتایج نشان داد استفاده از مدل شبکه عصبی موجک می‏تواند در زمینه تخمین خشکسالی موثر باشد که در نوبه خود برای تسهیل توسعه و پیاده سازی استراتژی های مدیریتی جهت جلوگیری از ایجاد خشکسالی مفید است.

کلیدواژه‌ها


عنوان مقاله [English]

Estimation Drought in Lorestan using Intelligent Networks

نویسندگان [English]

  • hassan Torabipodeh 1
  • Reza Dehghani 2
  • Saeed Rostami 3
1 Associate Professor of Water Engineering, Lorestan University
2 PHD student water structures, Lorestan University
3 PHD student water structures, Lorestan University
چکیده [English]

Drought is a natural and irreversible phenomenon that results from a reduction in rainfall over a given period of time. This phenomenon begins slowly and its impact gradually and over a relatively long period of time appears in different sectors, such as water resources, agriculture, the environment, and so on. Therefore, it is difficult to determine precisely the time of the onset and end of this phenomenon, due to the nature of the drought, it is difficult to detect the beginning and the end of the drought. Drought prediction in water resource systems plays an important role in reducing drought damage. Traditionally, in the last few decades, drought has been widely used to predict fit and mathematical models. For prediction of drought, a variety of approaches have been introduced in hydrology, in which intelligent models are the most important ones. In this study, monthly rainfall data of Nahrabad, Alshtar, Dorood and Boroujerd stations in Lorestan province were used to evaluate the accuracy of models in drought prediction. For modeling, wavelet network and artificial neural network models were used and the results were compared to each other for the accuracy of the studied models.
Materials and methods:In this research, four rain-impact stations of Nurabad, Alshatr, Dorood and Borujerd in Lorestan province were selected as the study area and drought analysis was carried out using SPI standard rainfall index at a 12-month time scale at these stations. . For this purpose, rainfall parameter was selected on monthly basis during the statistical period (1394-1374) as input and standard rainfall index as the output parameter of the models.
Wavenet called wavelet-based neural network which combined with wavelet theory and neural networks have been created.It also have supportive of the benefits and features of neural networks and charm and flexibility and strong mathematical foundations and analysis of multi-scale wavelet . a combination of wavelet theory with neural network concepts to the creation of wavelet neural network and feedforward neural shock  can be a good alternative for estimating approximate nonlinear functions .Feedforward neural network with sigmoid activation function is in the hidden layer While at the nerve shocked wavelet ,wavelet functions as activation function of hidden layer feedforward networks are considered, In both these networks and scale wavelet transformation parameters are optimized with their weight.
Artificial neural networks inspired by the brain's information processing systems, design and emerged intoTo help the learning process and with the use of processors called neurons trying to understand the inherent relationships between data mapping between input space and optimal space. Hidden layer or layers, the information received from the input layer and output layer are the processing and disposal.Based on the artificial neural network structure, its major features high processing speed, the ability to learn the pattern,The ability to extend the model after learning, flexibility against unwanted errorsNo disruption to   error on the part of the connection due to weight distribution network. The first practical application of synthetic networks with the introduction of Multilayer Perceptron network wasConsultants. for training this network back propagation algorithm is used.The basis of this algorithm is based on error correction learning ruleThat  consists of two main routes.By adjusting the parameters in the MLP model error signal and input signal occurs.Determine the number of layers and neurons is the most important issues in simulation with artificial neural network.
The criteria of correlation coefficient, root mean square error and of mean absolute error were used to evaluate and performance compare of models.
Results and Discussion: The results showed that both models have a good ability to estimate the standard rainfall index, but in terms of accuracy, the wavelet neural network model has shown better performance than artificial neural network.
The results also showed that the wavelet neural network model has less error than the artificial neural network, and this model (wavelet neural network) has shown an acceptable accuracy in estimating most of the values. On the other hand, the results of the drought index test showed that in both models, the Drood station is more consistent with observational values.
Conclusion:Overall, the results showed that the use of wavelet neural network model can be effective in drought estimation, which in turn is useful for facilitating the development and implementation of management strategies to prevent drought.

کلیدواژه‌ها [English]

  • Precipitation
  • Drought
  • Standardized precipitation index
  • Wavelet Neural Network

 

Afkhami, H., M. Ekhtesasi and M. Mohammadi, 2014, Effect of input variables preprocessing in SPI(Standardized Precipitation Index) prediction using artificial neural network and wavelet transformation. Iranian Journal of Range and Desert Research, Vol.22, No.3, pp.570-582
Barua, S., B.J.C. Perera, A.M. NG and D.H. Tran, 2010, Drought forecasting using an aggregated drought index and artificial neural networks. Journal of water and climate change, Vol. 1, No.3, pp. 206  193 .
Belayneh, A., J. Adamowski, B. Khalil and B.  Zielinski, 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, Vol.508, No.4, pp.418-429.
Djerbouai, M and D. Gamane, 2016, Drought Forecasting Using Neural Networks, Wavelet Neural Networks, and Stochastic Models: Case of the Algerois Basin in North Algeria. Water Resources Management: An International Journal, Published for the European Water Resources Association (EWRA), Vol.30, No.7, pp.2445-2464
Edwards, D.C, 1997, Characteristics of 20th century drought in the United States at multiple time scales. Climatology Report Number 97-2, Colorado State University, Fort Collins, Colorado
Gaye, O., O. Yildiz and A. Duvan, 2015,  A Drought Analysis Of Sivas Using The Standardized Precipitation Index (SPI) Method And Drought Estimation With The Artificial Neural Networks. International Journal of Advances in Mechanical and Civil Engineering, Vol.2, No.5, pp. 1-7
Golabi, M. R., F. Radmanesh and A. M. Akhondali, 2012, An Investigation of Artificial Neural Network and Time Series Performance in the Index Standard Precipitation Drought Modeling (Case Study: Selected Stations of Khuzestan Province). Arid Biome Scientific and Research Journal , Vol.3, No.1, pp.82-87
Jalalkamali, A., M. Moradi and M. Moradi, 2015, Application of several artificial intelligence models and ARIMAX model for forecasting drought using the Standardized Precipitation Index. International Journal of Environmental Science and Technology.  Vol.4, No.12, pp.1201-1210
Kaveh, A., and A. Iran Manesh, 2005, Artificial Neural Networks in Structural Optimization, Third Edition. Publications Building and Housing Research Center.
Kisi, O., M. Karahan and Z. Sen,  2006, River suspended sediment modeling using fuzzy logic approach .Hydrol Process, Vol.20, No.2, pp. 4351-4362.
Maca, P and P. Pech, 2016, Forecasting SPEI and SPI Drought Indices Using the Integrated Artificial Neural Networks. Computational Intelligence and Neuroscience, Vol.24, No.3, pp. 40-57
McKee, T.B., N.J. Doesken and J. Kleist, 1993, The Relationship of Drought Frequency and Duration to Time Scales. Eighth Conference on Applied Climatology, January 17e22, Anaheim, California, 179-184.
Mishra, A.K., V.R. Desai and P. Singh, 2007, Drought Forecasting Using a Hybrid Stochastic and Neural Network Model. Journal of Hydrologic Engineering, Vol.12, No.6, pp. 626-638
Nagy, H.,  K. Watanabe and M. Hirano, 2002, Prediction of sediment load concentration in rivers using artificial neural network model. Journal of Hydraulics Engineering, Vol.128, No.4, pp.558-559.
Nasiri, M. 2009, Drought prediction using Artificial Neural Networks Case Study of Shiraz Synoptic Station. Master's Degree in Climatology, Sistan and Baluchestan University
Negaresh, H and M. Armesh, 2010, Drought Forecasting in Khash City by Using Neural Network Model. Arid Regions Geographic Studies, Vol.2, No.6, pp.33-50
Nourani, V., M.T. Alami and M.H. Aminfar,  2009, A combined neural-wavelet model for prediction of Ligvanchai watershed precipitation. Engineering Applications of Artificial Intelligence, Vol. 22, No.2, pp.466–472.
Nourani, V., O. Kisi and M. Komasi, 2011. Two hybrid artificial intelligence approaches for modeling rainfall–runoff process. Journal of Hydrology, Vol.402, No. 1–2, pp. 41–59.
Safshekan, F., Pir Moradian, N., and R. A.  Sharifian, 2011, Simulation of rainfall-runoff hydrograph according to the time pattern of rainfall and the use of artificial neural network in the famous Kasilian Basin. Engineering Sciences Iranian Journal of Watershed Management, Vol.5, No.15, pp. 1-10
Shin, S., D. Kyung, S. Lee, J. Taik & Kim and J. Hyun, 2005, An application of support vector machines in bankruptcy prediction model. Expert Systems with Applications, Vol. 28, No.1, pp. 127-135.
Tokar, A and P. Johnson,1999, Rainfall-Runoff Modeling Using Artificial Neural Networks. J  Hydrol. Eng, Vol.4, No.3, pp.232-239
Vapnik, V.N, 1988, Statistical Learning Theory. Wiley, New York.
Wang, D., A.A. Safavi and J.A. Romagnoli, 2000,  Wavelet-based adaptive robust M-estimator for non-linear system identification. AIChE Journal, Vol. 46, No.8, pp.1607-1615.
Zhu, Y.M., X.X. Lu and Y. Zhou, 2007 Suspended sediment flux modeling with artificial neural network: An example of the Longchuanjian River in the Upper Yangtze Catchment. Geomorphology, Vol.84, No.1, pp.111-125.
Zulifqar,  A., I. Hussain, M. Faisal,  H. Mamona Nazir, T. Hussain, M. Yosafshad, A.M. Shoukry and S. Gani, S, 2017, Forecasting Drought Using Multilayer Perceptron Artificial Neural Network Model. Advances in Meteorology, Vol.25, No.1, pp.1-10