Journal of New Researches in Mathematics
,
Issue34,Year,
Winter
2022
یکی از مهمترین و بنیادیترین عامل حیات موجودات زنده آب است، لذا آلودگی آب ها، یک معضل بزرگ زیست محیطی محسوب می شود و جلوگیری از آلودگی آب ها و ارائه روش های هوشمند برای تصفیه آب ها بسیار مهم و مورد توجه است. تجهیز علوم مهندسی به ابزارهای هوشمند و هوش مصنوعی در تشخیص کیف More
یکی از مهمترین و بنیادیترین عامل حیات موجودات زنده آب است، لذا آلودگی آب ها، یک معضل بزرگ زیست محیطی محسوب می شود و جلوگیری از آلودگی آب ها و ارائه روش های هوشمند برای تصفیه آب ها بسیار مهم و مورد توجه است. تجهیز علوم مهندسی به ابزارهای هوشمند و هوش مصنوعی در تشخیص کیفیت تصفیه فاضلاب ها می تواند اشتباهات افراد خبره و خسارت های مالی ناشی از آن را کاهش دهد. تا کنون از روش های مختلفی برای تصفیه پساب های صنعتی استفاده شده است. اما با توجه به وقت گیر بودن و هزینه بالای این روش ها، استفاده از روش های کم هزینه و دقیق همواره مورد نیاز می باشد. دراین مقاله یک روش هوشمند ساده و ترکیبی بر پایه شبکه عصبی مصنوعی و روش آماری رگرسیون لجستیک ، جهت مدلسازی پیش بینی کیفیت خروجی سیستم های تصفیه فاضلاب ارائه می شود. سیستم هوشمند ارائه شده نقش مهمی در بررسی کیفیت تصفیه فاضلاب ها داشته و برای محققان هوش مصنوعی و مهندسین محیط زیست قابل استفاده می باشد.مقایسه نتایج پیش بینی شده توسط مدل شبکه عصبی ساده و مدل ترکیبی طراحی شده با پایه شبکه عصبی و رگرسیون لجستیک، نشان داد که روش پیشنهادی در این تحقیق یک روش ارزشمند برای پیش بینی کیفیت خروجی حاصل از تصفیه فاضلاب ها با بیشترین بازده وکمترین خطا می باشد.
Manuscript profile
International Journal of Industrial Mathematics
,
Issue5,Year,
Summer
2023
In current research, an architecture of hybrid articial neural networks has been employed to solve a special kind of fuzzy systems. The proposed four-layer fuzzied recurrent network can approximate real solution of the present fuzzy system to any desired degree of acc More
In current research, an architecture of hybrid articial neural networks has been employed to solve a special kind of fuzzy systems. The proposed four-layer fuzzied recurrent network can approximate real solution of the present fuzzy system to any desired degree of accuracy. To do this, a back-propagation learning rule based on the gradient descent method is designed to estimate the unknowns. Finally, some numerical experiments with comparison are presented to show the effectiveness of the recurrent back-propagation method.In current research, an architecture of hybrid articial neural networks has been employed to solve a special kind of fuzzy systems. The proposed four-layer fuzzied recurrent network can approximate real solution of the present fuzzy system to any desired degree of accuracy. To do this, a back-propagation learning rule based on the gradient descent method is designed to estimate the unknowns. Finally, some numerical experiments with comparison are presented to show the effectiveness of the recurrent back-propagation method.
Manuscript profile
International Journal of Industrial Mathematics
,
Issue1,Year,
Winter
2023
Integro-differential equations arise in various physical and biological problems. In this paper, a new iterative technique for solving linear Volterra-Fredholm integro-differential equation (VFIDE) has been introduced. The method is discussed in details and it is illust More
Integro-differential equations arise in various physical and biological problems. In this paper, a new iterative technique for solving linear Volterra-Fredholm integro-differential equation (VFIDE) has been introduced. The method is discussed in details and it is illustrated by solving some numerical examples. The approximate solution is most easily produced iteratively via the recurrence relation. Results are compared with the exact solutions, which reveal that new iteration method is very effective and convenient.
Manuscript profile
International Journal of Industrial Mathematics
,
Issue5,Year,
Autumn
2013
This paper intends to offer a new iterative method based on articial neural networks for finding solution of a fuzzy equations system. Our proposed fuzzied neural network is a ve-layer feedback neural network that corresponding connection weights to output layer are More
This paper intends to offer a new iterative method based on articial neural networks for finding solution of a fuzzy equations system. Our proposed fuzzied neural network is a ve-layer feedback neural network that corresponding connection weights to output layer are fuzzy numbers. This architecture of articial neural networks, can get a real input vector and calculates its corresponding fuzzy output. In order to nd the approximate solution of the fuzzy system that supposedly has a real solution, rst a cost function is dened for the level sets of the fuzzy network and target output. Then a learning algorithm based on the gradient descent method is used to adjust the crisp input signals. The present method is illustrated by several examples with computer simulations.
Manuscript profile
Sanad
Sanad is a platform for managing Azad University publications