ارزیابی و مقایسۀ سیستم استنتاج فازی- عصبی و شبکۀ عصبی مصنوعی پرسپترون چند لایه در برآورد هدایت هیدرولیکی اشباع خاک با استفاده از بافت خاک (مطالعۀ موردی: شبکۀ آبیاری دشت فتحعلی مغان)

نوع مقاله: مقاله پژوهشی

نویسندگان

1 دانشگاه محقق اردبیلی

2 دانشکده فنی وحرفه ای سما،دانشگاه آزاد اردبیل،اردبیل،ایران

چکیده

اندازه­گیری مستقیم ویژگی‌های هیدرولیکی خاک وقت­گیر و پر هزینه است اما می‌توان این ویژگی‌هارا با بهره­گیری از داده­ های زودیاف مثل بافت خاک، جرم مخصوص ظاهری و با استفاده از روش‌هایی چون توابع انتقالی و سیستم استنتاج فازی- عصبی نیز به دست آورد. در این تحقیق برای برآورد هدایت هیدرولیکی اشباع خاک، ازمدلشبکۀ عصبی مصنوعی و سیستم­استنتاجفازی-عصبیاستفاده شد. ورودی­های مدل، شامل درصد رس، سیلت و شن بود. معماری شبکۀ عصبی دارای 3 نرون در لایۀ ورودی، 11 نرون در لایۀ پنهان با تابع انتقال تانژانت سیگموئید و یک نرون در لایۀ خروجی با تابع انتقال خطی با 1000 تکرار بود و در تمام شبکه­ از سرعت یادگیری و مومنتم مساوی با 3/0 استفاده شد. سیستم استنتاج فازی- عصبی دارای 27 قانون است و برای تابع عضویت متغیرهای ورودی از تابع گوسین استفاده شد. همچنین، برای بهینه سازی سیستم استنتاج فازی- عصبی از روش هیبرید استفاده شد. برای ارزیابی عملکرد مدل از پارامترهای مجذور میانگین مربعات خطا (سانتی‌متر بر روز)، درصد خطای نسبی، میانگین خطای مطلق (سانتی‌متر بر روز)، ضریب جرم باقیمانده، راندمان مدل و ضریب تبیین استفاده شد که برای مدل فازی- عصبی به ترتیب 032/0، 627/0، 18/0، 0000023/0-، 999/0 و 997/0 به دست آمد. برای شبکۀ عصبی مصنوعی نیز با الگوریتم آموزشی لونبرگ مارکوت در تخمین هدایت هیدرولیکی اشباع خاک این مقادیر به ترتیب 22/1، 44/1، 21/1، 00015/0-، 997/0 و 99/0 به دست آمد. نتایج تحقیق نشان می‌دهد که سیستم استنتاج فازی- عصبی نسبت به شبکۀ عصبی مصنوعی دقیق­تر است و نسبت به داده­های اندازه­گیری شده نتایجی نزدیکتر ارائه می‌دهد.

کلیدواژه‌ها


عنوان مقاله [English]

Adaptive Neuro Fuzzy Inference System and Multilayer Perceptron Neural Networks to Estimate Saturated Hydraulic Conductivity by Soil Texture A Case Study for Fath-Ali Irrigation Network in Moghan Plain

چکیده [English]

Direct measurement of soil hydraulic conductivity is time-consuming and expensive. Direct measurement of soil hydraulic properties can be replaced by simple measurement of properties such as soil texture and bulk density using transfer functions and an adaptive neuro fuzzy inference system (ANFIS). The present study used ANFIS and neural network models to estimate saturated soil hydraulic conductivity. The model inputs included percentage of silt, clay, and sand. The architecture for this network contained 3 neurons in the input layer and 11 neurons in the hidden layer using the tangent sigmoid transfer function, and an output layer of neurons with a linear transfer function and 1000 iterations. In all networks, the learning rate and momentum was 0.3. The neuro fuzzy inference system had 27 rules, a Gaussian membership function was used for input data, and a hybrid method was used to optimize the ANFIS model. The root mean square error (mmd-1), percentage of relative error (ε), mean absolute error (cmd-1), coefficient of residual mass, efficiency, and coefficient of determination were used to evaluate the performance of the model. For the ANFIS model, these values were 0.032, 0.62%, 0.18, -0.0000023, 0.999, and 0.997, respectively. The values for the Levenberg-Marquardt training algorithm were 1.22, 1.44%, 1.21, -0.00015, 0.997, and 0.99, respectively. Performance evaluation of the models showed that the ANFIS model predicted soil hydraulic conductivity with greater accuracy than did the neural network and the results of this method were closer to actual measurement results.

کلیدواژه‌ها [English]

  • ANFIS
  • ANN
  • Estimation
  • Saturated hydraulic conductivity
  • Soil Gradation

Agyare, W. A., Park, S. J. and Vlek, P. L. G. 2007. Artificial neural network estimation of saturated hydraulic conductivity. Vadose Zone J. 6(2): 423-431.

 

Amir-Abedi, H. 2012. Estimation soil hydraulic properties of Ardabil plain using artificial neural networks and regression methods. M. Sc. Thesis. Faculty of Agricultural Technology and Natural Resources. University of Mohaghegh Ardabili. (in Persian)

 

Amjadi, N. 2002. Introduction to Intelligent Systems. Semnan University Press. (in Persian)

 

Anon. 2008. MATLAB software help manual. Version7.6.0.2008b. The Math Works, Inc.

 

Aqil, M., Kita, I., Yano, A. and Nishiyama, S. 2007. A comparative study of artificial neural networks
and neuro-fuzzy in continuous modeling of the daily and hourly behaviour of runoff. J. Hydrol.
337(1-2): 22-34.

 

Demuth, H. and Beale, M. 2000. Neural network toolbox user Guide. Copyright 1992-2002. By The Math Works, Inc. Version4.

 

Finol, J., Guo, Y. K. and Jing, X. D. 2001. A rule based fuzzy model for the prediction of petrophysical rock parameters. J. Petrol. Sci. Eng. 29, 97-113.

 

Firoozi, S., Sheikhdavoodi, M. J. and Sami, M. 2014. Evaluation the ability of different artificial
intelligence-based modeling techniques in prediction of yield using energy inputs data of farms. J. life Sci. Biomed. 4(3): 214-220.

 

Givi, J., Prasherb, S. O. and Patel, R. M. 2004. Evaluation of pedotransfer functions in predicting the soil water contents at field capacity and wilting point. Agri. Water Manage. 70, 83-96.

 

Gokceoglu, C. 2002. A fuzzy triangular chart to predict the uniaxial compressive strength of the Ankara agglomerates from their petrographic composition. Eng. Geol. 66, 39-51.

 

Heusher, S. A., Brandt, C. C. and Jardin, P. M. 2005. Using soil physical and chemical properties to estimate bulk density. Soil Sci. Soc. Am. J. 69, 51-56.

 

Jacovides, C. P. 1997. Reply to comment on Statistical procedures for the evaluation of evapotranspiration models. Agr. Water Manage. 3, 95-97.

 

Jang, J. S. R. 1993. ANFIS: Adaptive Network Based Fuzzy Inference System. IEEE Trans. Syst. Man Cybern. 23 (3): 665-685.

 

Jang, J. S. R. and Sun, C. T. 1997. Neuro-fuzzy modeling and control. Proceedings of the IEEE. 83(3):
378-406.

 

Karakus, M. and Tutmez. B. 2006. Fuzzy and multiple regression modeling for evaluation of intact rock strength based on point load, Schmidt hammer and sonic velocity, Rock Mech. Rock Eng.
39(1): 45-57.

 

Karatalopoulos, S. V. 2000. Understanding Neural Networks and Fuzzy Logic-Basic Concepts and Applications. Wiley-IEEE Press. New-Delhi. India.

 

Kianpur-Kalkhajeh, U., Rezaie-Arshad, R., Amerikhah, H. and Sami. M. 2010. Multiple linear regression, artificial neural network (MLP, RBF) and ANFIS models for modeling the saturated hydraulic conductivity of tropical region soils (case study: Khuzestan province: sothwest Iran). Int. J. Agr. Res and Rev. 2(3): 255-265.

 

Lippman, R. P. 1987. An introduction to computing with neural nets. ASSP Magazine, IEEE. 4(2):4-22.

 

Luk, K. C., Ball, J. E. and Sharma. A. 2000. A study of optimal model lag and spatial inputs to artificial neural network for rainfall forecasting. J. Hydrol. 227(1-4): 56-65.

 

Marshal, T. J. 1958. A relationship between permeability and size distribution of pores. Soil Sci. 9, 1-8.

 

Menhaj, M. B. 1998. Principle of Neural Networks. First edition. Professor Hesabi Publishing Center. Tehran. Iran.

 

Merdun, H., Ozer, C., Meral, R. and Apan, M. 2006. Comparison of Artificial Neural Network and regression pedotransfer functions for prediction of soil water retention and saturated hydraulic conductivity. Soil Till. Res. 90(1-2): 108-116.

 

Mohammadi, J. 2002. Testing an artificial neural network for predicting soil water retention characteristics from soil physical and chemical properties. 17th WCSS. Thailand. Paper No. 221.

 

Mohammadi, J. and Taheri, M. 2005. Estimation of pedotransfer function using fuzzy regression. J. Agric. Sci. Technol. 9(2): 51-60. (in Persian)

 

Naderloo, L., Alimardani, R., Omid, M., Sarmadian, F., Javadikia, P., Yaser Torabi, M. and Alimardani. F. 2012. Application of ANFIS to predict crop yield based on different energy inputs. Measurement. 45, 1406-1413.

 

Nosrati-Karizak, F., Movahhedi-Naieni, S. A. R., Hezarjaribi, A., Roshani, Gh. A. and Dehghani, A. H. 2012. Using artificial neural networks to estimate the saturated hydraulic conductivity of readily available soil properties. Soil Manage. Sustain. Prod. 2(1): 95-110.

 

Rahimi-Ajdadi, F. and Abbaspour-Gilandeh. Y. 2011. Artificial neural network and stepwise multiple range regression methods for prediction of tractor fuel consumption. Measurement. 44(10): 2104-2111.

 

Rawls, W. J. 2004. Pedotransfer functions for the United States. Dev. Soil Sci. 30, 437-447.

 

Rehman, S. and Mohandes, M. 2008. Artificial neural network estimation of global solar radiation using air temperature and relative humidity. Energ. Policy. 36, 571-576.

 

Rezaie-Arshad. R. 2010. Expansion pedo transfer functions of soil hydraulic conductivity with artificial neural networks and statistical regression methods for Khouzestan province soils. M.S Thesis. Faculty of Agricultural Science. University of Shahid Chamran. (in Persian)

 

Salazar, O., Wesstrom, I. and Joel. A. 2008. Evaluation of Drainmod using saturated hydraulic conductivity estimated by a pedotransfer function model. J. Agr. Water Manage. 95, 1135-1143.

 

Sarmadian, F., Taghizadeh-Mehrjerdi, R., Mohammad-Askari, H. and Akbarzadeh, A. 2010. Comparison of nero fuzzy, neural network and multiple regression methods to predicting some soil properties (Case Study: Golestan province). Iranian J. Soil Water Res. 41(1): 211-220.

 

Schaap, M. G. and Leij, F. J. 1998. Using neural networkes to predict soil water retention and soil hydraulic conductivity. Soil Till. Res. 47, 37-42.

 

Soltani, S. and Morid, S. 2002. Comparison of Conceptual model with artificial neural network to simulate rainfall-runoff process. Proceeding of the 6th International River Engineering Conference. Shahid Chamran University. Ahvaz. Iran. (in Persian)

 

Tamari, S., Wosten, J. H. M. and Ruz-suarez, J. C. 1996. Testing an artificial neural network for predicting soil hydraulic conductivity. Soil Sci. Soc. Am. J. 60, 1732-1741.

 

Vos, B. D., Meirvenne, M. V., Quataert, P., Deckers, J. and Muys, B. 2005. Predictive quality of pedotransfer functions for estimating bulk density of forest soils. Soil Sci. Soc. Am. J. 69, 500-510.

 

Wosten, J. H. M., Lilly, A. Nemes, A. and Bas, C. 1999. Development and use of a database of hydraulicproperties of European soils. Geoderma. 90, 169-185.

 

Yilmaz, I. and Kaynar, O. 2011. Multiple regression, ANN (RBF, MLP) and ANFIS models for prediction of swell potential of clayey soils. Expert Syst. Appl. 38(5): 5958-5966.

 

Schaap, M. G. and Leij, F. J. 1998. Using neural networkes to predict soil water retention and soil hydraulic conductivity. Soil Till. Res. 47, 37-42.

 

Soltani, S. and Morid, S. 2002. Comparison of Conceptual model with artificial neural network to simulate rainfall-runoff process. Proceeding of the 6th International River Engineering Conference. Shahid Chamran University. Ahvaz. Iran. (in Persian)

 

Tamari, S., Wosten, J. H. M. and Ruz-suarez, J. C. 1996. Testing an artificial neural network for predicting soil hydraulic conductivity. Soil Sci. Soc. Am. J. 60, 1732-1741.

 

Vos, B. D., Meirvenne, M. V., Quataert, P., Deckers, J. and Muys, B. 2005. Predictive quality of pedotransfer functions for estimating bulk density of forest soils. Soil Sci. Soc. Am. J. 69, 500-510.

 

Wosten, J. H. M., Lilly, A. Nemes, A. and Bas, C. 1999. Development and use of a database of hydraulicproperties of European soils. Geoderma. 90, 169-185.

 

Yilmaz, I. and Kaynar, O. 2011. Multiple regression, ANN (RBF, MLP) and ANFIS models for prediction of swell potential of clayey soils. Expert Syst. Appl. 38(5): 5958-5966.