تخمین ارتفاع امواج ناشی از باد با استفاده از سیستم استنتاج تطبیقی عصبی- فازی، درخت تصمیم و روش های تجربی در بندر بوشهر

نوع مقاله : مقاله پژوهشی

نویسندگان

1 دانشگاه علوم و فنون دریایی خرمشهر

2 استادیارگروه مهندسی سازه‌های دریایی، دانشگاه علوم و فنون دریایی خرمشهر

3 استادیارگروه مهندسی سازه‌های دریایی، دانشگاه علوم و فنون دریایی خرمشهر.

4 استادیار گروه الکترونیک و مخابرات دریایی، دانشگاه علوم و فنون دریایی خرمشهر.

چکیده

اطلاعات حاصل از پیش­بینی امواج بعنوان اساس طراحی­های مهندسی دریا، نقش مهمی در ساخت، نگهداری و مدیریت پروژه­های عمرانی ساحلی و فراساحلی، دارد. امواج ناشی از باد، به دلیل دارا بودن بیش‌ترین انرژی و فراوانی، از مهم‌ترین امواج در دریا محسوب می­شوند. در تحقیق حاضر، از سیستم استنتاج تطبیقی عصبی- فازی، برای پیش­بینی خصوصیات امواج  استفاده شده است. منطقه مورد مطالعه، بندر بوشهر و پیش بینی­ها برای بازه زمانی سال 2008 تا سال 2010 در نظر گرفته شده است. نتایج نشان داد در پیش­بینی ارتفاع امواج ناشی از باد، دقت مدل سیستم عصبی- فازی در مقایسه با  درخت تصمیم، در مدت زمان سه و شش ساعته مشابه بدست آمد. به­طوری­که در مدت زمان سه ساعته ضریب همبستگی 86 و 83 درصد برای پیش بینی سه ساعته و 78 و 74 درصد برای پیش­بینی شش ساعته دارای دقتی بالاتر از مدل‌های تجربی می‌باشند مقایسه­ی نتایج پیش‌بینی با روش­های تجربی نشان داد، که روش SMB با ضریب همبستگی بیشتر و درصد خطای کمتر نسبت به بقیه روش­های تجربی، در برآورد ارتفاع امواج بیش‌ترین دقت را در محاسبات داشت . علاوه بر آن روش SPM با ضریب همبستگی 72 و 63 درصد برای پیش­بینی  سه و شش ساعته، نامناسب‌ترین روش برای تعیین ارتفاع موج بوده و ارتفاع موج را کمتر از شرایط واقعی پیش‌بینی می‌کند.   

کلیدواژه‌ها

موضوعات


عنوان مقاله [English]

Height prediction of the Wind – Induced Waves using Adaptive – Network –Based Fuzzy Inference System, Decision Tree and Empirical Methods ( Case study: Bushehr Port) Bushehr Port

نویسندگان [English]

  • Liyaghat Bozorgzadeh 1
  • Morteza Bakhtiari 2
  • Nima ShahniKaramzadeh 3
  • Mohammad Esmaeel Doust 4
1 Master of science coastal Engineering , Khorramshhar university of Marine Sciecne and Technology.
2 Assistant Professor, Department of Maine Structures , Khorramshhar University of Marine Sciecne and Technology
3 Assistant Professor, Department of Maine Structures , Khorramshhar university of Marine sciecne and Technology.
4 Assistant Professor , Department of marine Electronic and communication Engineering , Khorramshhar University of Marine Sciecne and Technology.
چکیده [English]

 Derived Information from the predicted waves will be the basis for engineering design and plays an important role in the construction, maintenance and management of coastal and offshore construction projects. The wind-induced-waves  due to having high energy and frequency are of great importance in the sea. In this research,  the adaptive network based fuzzy inference system was selected to predict the waves’ characteristics.. Study About prediction of wave Characteristics is carried out by several researchers. Zarghani et al  (2006) studied the characteristics of wind-induced waves at the coast of Khark island They used the SPM model to conduct the survey. The results obtained from the model showed that the highest wave was 3.7 m with rotation period of  6.8 seconds. Taleghani and Amirteimori (2008) used the field data of the Caspian Sea waves measured by a waveguide buoy in an artificial neural network. Finally, the comparison of real data measured by the measurement systems with the results of the neural network is a good match that indicates the accuracy and speed of the method used in the short term. Kamranzad and Etemad Shahidi (2011) studied the prediction of wind-induced waves in Assaluyeh using the SWAN numerical model. The validation results of the model showed that the results of calibration model have a good accuracy. The calculated error indices in validation periods also showed the acceptable accuracy of SWAN modeling. Therefore, the constructed model had the ability to predict wave height and period of time in Assalouyeh. In order to do present research, the Adaptive Nero-Fuzzy Inference System as one of the soft computing methods is used. The study area chosen Bushehr and forecasts for the period December 2008 to 2010 is considered. The results showed that, in prediction height of wave induced by wind, accuracy of the Adaptive Nero-Fuzzy Inference System and decision tree are same and are higher than the empirical methods. The comparison results of prediction from empirical methods showed that among the experimental methods, SMB method with high correlation coefficient and less error percent than the other experimental methods has higher accuracy in estimating height of waves in addition the SPM method is not suitable method for determining the height of waves and under estimated the real values.

کلیدواژه‌ها [English]

  • Wind – induced waves
  • Waves height
  • Adaptive – Network –Based Fuzzy Inference System
  • Decision tree
  • Bushehr port
1-     Anonymous 2002. Coastal Engineering Manual. US Army Corps of Engineers. Part 2. Chapter 1.
 
2-    Bishop, C. T .1983.Comparison of Manual wave Prediction Models. Journal of Waterway Port Coastal & Ocean Engineering, ASCE, 109.pp. 1-17.
 
3-    Chegini, F., Chegini, V. and Taebi, S.2008. Evaluating and comparing semi-experimental and numerical methods in predicting the characteristics of the waves of Amirabad and Bushehr ports. Journal of Marine Engineering,4(7).pp.(In Persian).
 
4-    Deo, M.C., Jha, A., Chaphekar, A.S. and Ravikant ,K. 2001. Neural network for wave forecasting Journal of Waterway Port Coastal & Ocean Engineering .28.pp. 889–898.
 
5-    Derakhshan,Sh., Mostafa Gharebaghi,A. and Chanaghloo, M.2004. Projection of sea waves with experimental methods in Bushehr area. In 1th national Civil Engineering Conference, Sharif University of Technology, Iran. (In Persian).
 
6-    Donelan, M.A.1980. Similarity theory applied to the forecasting of wave heights, periods and directions. In: Proceeding of the Canadian Coastal Conference, National Research Council of Canada.pp. 47–61.
 
7-     Etemad-Shahidi, A., Mahjoobi, J .2009. Comparison between M5– model tree and neural networks for prediction of significant wave height in Lake Superior, Journal of Waterway Port Coastal & Ocean Engineering.36.pp. 1175–1181.
 
8-    Kamranzad, B. and Etemad Shahidi,A.2011. Prediction of wind waves in Assaluyeh using numerical model. Thesis. Iran University of Science and Technology. 120p.(In Persian).
 
9-    Kamranzad, B., Etemad-Shahidi, A. and Kazeminezhad, M.H. 2011. Wave height forecasting in Dayyer, the Persian Gulf. Journal of Ocean Engineering. 38 (2011).pp. 248–255.
 
10- Mavedatnia,H., Bakhtiari,M., Bahrami,H. and Behdarvandi Asgar,M.2014. Prediction of Wave Characteristics in Imam Khomeini Port Using Artificial Neural Network Software. Thesis, Khorramshar University of Marine Science and Technology.120p.
 
11- Shahidi,A.E.,Moeini,M.H. and Kazeminezhad,M.H.,2010.Wind –induced Waves: theory, prediction methods and models. Iran University of Science and Technology. .(In Persian).
 
12- Taleghani, M.and Amir Teimori,A. 2008. Projection of Caspian Sea Waves Using Artificial Neural Network. Journal of Applied Mathematics.,5(18),Pp.39-47.(In Persian).
 
13- Tsai, C.P., Lin, C., Shen, J.N., 2002. Neural network for wave forecasting among multi- stations. Journal of Ocean Engineering 29.pp.1683–1695.
 
14- Zamani, A., Solomatine, D., Azimian, A. and Heemink ,A. 2008. Learning from data for wind wave forecasting. Journal of Waterway Port Coastal & Ocean Engineering .35.pp. 953–962.
 
15- Zarghani.2006. Prediction of distant waves from the island of Khark. Thesis, Science and Reacherch Branch Islamic University of Tehran, Iran. 131p. (In Persian).