بهره برداری بهینه منابع آب در زمان واقعی بر اساس الگوریتم NSGA-II و ماشین های بردار پشتیبان (مطالعه موردی: سد گاوشان)

نوع مقاله: مقاله پژوهشی

نویسندگان

استادیار گروه مهندسی آب دانشکده علوم و مهندسی کشاورزی، دانشگاه رازی.

چکیده

در این تحقیق از ترکیب الگوریتم چندهدفه NSGA-II و مدل شبیه­ساز WEAP برای استخراج سیاست­های بهینه بهره­برداری از مخزن در قالب بهینه­سازی معین استفاده شد. طوری­که در آن، هدف اول، حداکثر نمودن اطمینان­پذیری تأمین نیازها در مقابل هدف دوم یعنی حداقل نمودن میزان تخطی ناشی از عدم تأمین نیازها و تخطی از ظرفیت مخزن در طول دوره بهره­برداری قرار گرفت. اما جواب­های بهینه یعنی مقدار رهاسازی از مخزن قابل تعمیم برای سایر ورودی­های محتمل به مخزن نیستند. در صورت تغییر جریان ورودی به مخازن جواب­های بهینه به­دست آمده کارایی نداشته و باید بهره­برداری از سیستم در قالب الگوریتم بهینه­ساز مجددا بهینه گردد. لذا برای حل این مشکل روش جدیدی بر اساس تلفیق روش ماشین بردار پشتیبان و الگوریتم NSGA-II برای بهره­­برداری بهینه از سیستم در زمان واقعی توسعه داده شد. نتایج نشان داد میانگین خطای قوانین بهینه مستخرج از ماشین­های بردار پشتیبان نسبت به خروجی الگوریتم NSGA-II در مرحله صحت­سنجی کمتر از 10 درصد است که نشان دهنده کارایی این روش در پیش­بینی الگوی بهینه منحنی فرمان سد در زمان واقعی است. در این ساختار می­توان بر اساس جریان ورودی به مخزن، حجم ذخیره آب در مخزن و تغییرات ذخیره مخزن (در ابتدای ماه) و نیازهای پایین­دست در ماه حاضر، مقدار رهاسازی بهینه را در زمان واقعی به­­دست آورد. روش مورد استفاده این قابلیت را داشته که با توجه به ورودی­های جدید جریان به سد، سریعا سیاست­های بهره برداری بهینه را به نحوی در اختیار قرار دهد که امکان مدیریت بهینه سیستم در زمان واقعی فراهم گردد.

کلیدواژه‌ها

موضوعات


عنوان مقاله [English]

Optimal Utilization of Water Resources in Real Time Based on NSGA-II Algorithms and Support Vector Machines (Case Study: Gavoshan Dam)

نویسندگان [English]

  • Arash Azari
  • Ali Arman
Assistant Professor, Department of Water Engineering, Razi University
چکیده [English]

Ample research has been done on optimization methods for the exploitation of reservoirs in the form of a specific optimization. In this type of studies, a certain series of flow is provided for the reservoir during the operation period and the release from the reservoir to downstream is optimized under such circumstances. They are certain drawbacks for these models. First, optimal solutions cannot be generalized for other possible inputs to the reservoir. Second, in the event of a change in the flow of input into the reservoirs, it is likely that the optimal solutions are not efficient and the operation of the system in the form of an optimization algorithm should be performed again. In such circumstances, one of the solutions is the use of intelligent methods such as support vector machines to apply the results of system optimization in real time. The main goal of this study is to integrate artificial intelligence methods such as support vector machine with NSGA-II optimization algorithm to convert it to real time. In this structure, contrary to the conventional design of the scheduling, in the event of a change in the flow of input, there is no need to perform re-optimization to understand the optimal coefficients. Instead, the extraction relationship of the support vector machine can be based on the input to the reservoir (at the beginning of the month), the volume of water storage in the tank (at the beginning of the month), reservoir storage changes and downstream needs in the current month to yield optimal release rates in real time.

کلیدواژه‌ها [English]

  • Artificial Intelligence
  • Optimization
  • On-time Operation
  • Metaheuristic Algorithm
  • WEAP
1- Bazargan-Lari, M.R., Kerachian, R., Sedghi, H., Fallahnia, M., Abed-Elmdoust, A. and Nikoo, M.R., 2011. Developing probabilistic operating rules for real-time conjunctive use of surface and groundwater resources: Application of support vector machines. J. of Water and Wastewater, 76, pp.54-69. (In Persian).

 

2- Blum, C., and Roli, A., 2003. Metaheuristics in combinational optimization: Overview and conceptual comparision, ACM Computing Surveys, 35(3), pp. 268-308.

 

3- Candela, J. Q. and Hansen, L. K., 2002. Time series prediction based on the Relevance Vector Machine with adaptive kernels, Acoustics, Speech, and Signal Processing (ICASSP), In IEEE International Conference on, Orlando, FL, USA

 

4- Chang, J. F., Chen, L., and Chang, C. L., 2005. Optimizing reservoir operating rule curves by genetic algorithms, Hydrological Processes, 19, pp. 2277-2289.

 

5- Deb, k., Pratap, A., Agarwal, S., and Meyarivan, T., 2002. A fast and elitist multi-objective genetic algorithm: NSGA-II, IEEE Trans Evolutionary Computing, Indian, 6(2), pp. 182–197.

 

6- Du, J., Liu, Y., Yu, Y. and Yan, W., 2017. A Prediction of Precipitation Data Based on Support Vector Machine and Particle Swarm Optimization (PSO-SVM) Algorithms, Algorithms, 10 (57), pp. 1-15.

 

7- Esmikhani, M., Safavi, H. and Yazdani-Poor, M., 2010. Integrated management of surface water and groundwater resources using the methods of suport vector machines and genetic algorithms, In 5th National Congress on Civil Engineering, Ferdowsi University of Mashhad, Mashhad, Iran.

 

8- Gen, M., and Cheng, R. W., 1997. Genetic Algorithms and Engineering Design. John Wiley & Sons, Inc.

 

9- Hojjati, A., FaridHoseini, A. R., Ghahreman, B. and Alizadeh, A., 2013. The comparison of the application of heuristic methods in optimization of multiobjective water resources systems, Water and Environment Engineering, 1(2), pp. 9-14. (In Persian).

 

10- Huang, M. and Tian, Y., 2015. Prediction of Groundwater Level for Sustainable Water Management in an Arid Basin Using Data driven Models, In International Conference on Sustainable Energy and Environmental Engineering (SEEE), Bangkok.

 

11- Jian, C. Qiang, H., and Min, W., 2005. Genetic algorithm for optimal dispatchin, Water Resource Planning and Management, 19, pp. 321-331.

 

12- Li, X., Sha, J., Li, Y. M. and Wang, Z. L., 2017. Comparison of hybrid models for daily stream flow prediction in a forested basin, Journal of Hydro informatics, Available Online 29 November 2017, 20(1):jh2017189, doi: 10.2166/hydro.2017.189.

 

13- Lin, J. Y., Cheng, C. T. and Chau, K. W., 2006. Using support vector machines for long-term discharge prediction, Hydrological Sciences–Journal–des Sciences Hydrologiques, 51(4), pp. 599-612.

 

14- Mirfendreski, G. and Mosavi, S.J., 2011. Supported vector machines for super-modeling in order to optimize the allocation of basin water resources, In 6th National Congress on Civil Engineering, Semnan University, Semnan, Iran.

 

15- Momtahen, Sh., and Dariane, A. B., 2007. Direct search approaches using genetic algorithms for optimization of water reservoir operating policies, Water Resource Planning and Management, ASCE, 133(3), pp. 202–209.

 

16- Nakhaei, M., Mohammadi, M. and Rezaei, M., 2014. Optimizing of Aquifer Withdrawal Numerical Model UsingGenetic Algorithm (Case Study: Uromiyeh Coastal Aquifer), Iran Water Resources Research, 10(2), pp. 94-97. (In Persian).

 

17- Nikoo, M. R., Karimi, A., Kerachian, R. and Bashi-Azghadi, S. N., 2013. Long-term Optimum Water and Waste Load Allocation Rules in River-ReservoirGroundwater Systems: Application of SVR Simulation Model, Water and Environment Engineering, 1(1), pp. 1-11. (In Persian).

 

18- Noori, R., Khakpour, A., Dehghani, M. and Farokhnia, A., 2010. Monthly Stream Flow Prediction Using Support Vector Machine Based on Principal Component Analysis, Journal of Water and Wastewater, 1, pp.118-129. (In Persian).

 

19- Schölkopf, B. and Smola, A., 2002. Learning with Kernels. MIT Press.

 

20- Shirzad, A., Tabesh, M. and Farmani. R. 2014. A Comparison between Performance of Support Vector Regression and Artificial Neural Network in Prediction of Pipe Burst Rate in Water Distribution Networks, KSCE Journal of Civil Engineering, 18(4), pp. 941-948.

 

21- Smola, A. J and Schölkopf, B., 1998. On a kernel–based method for pattern recognition, regression, approximation and operator inversion, Algorithmica, 22, pp. 211–231.

 

22- Tennant, D. L., 1976. Instream flow regimens for fish, wildlife, recreation and related environmental resources, Fisheries, 1(4), pp. 6-10.

 

23- Thissen, U., van Brakel, R., de Weijer, A. P., Melssen, W.J. and Buydens, L. M. C., 2003. Using support vector machines for time series prediction, Chemometrics and Intelligent Laboratory Systems, 69, pp. 35–49.

 

24- Vapnik, V., 1995. The nature of statistical learning theory. Springer-Verlag, NewYork.

 

25- Wardlaw, R., and Sharif, M., 1999. Evaluation of genetic algorithms for optimal reservoir system operation, Water Resource Planning and Management, ASCE. 125(1), pp. 25-33.

 

26- Xi, X. C., Poo, A. N. and S. K. Cho., 2007. Support vector regression model predictive control on a HVAC plant, Control Engineering Practice, 15, pp. 897–908.