Solving Engineering Optimization Problems with the Karush-Kuhn-Tucker Hopfield Neural Networks

Ganesan, T.; Elamvazuthi, I.; Vasant, P.
November 2011
International Review of Mechanical Engineering;Nov2011, Vol. 5 Issue 7, p1333
Academic Journal
The Karush-Kuhn-Tucker (KKT) approach is a well established classical method to solve non-linear programming (NLP) optimization problems. The aim of this work is to integrate the KKT method into the Hopfield Neural Networks (HNN) and hence create a new algorithm, the KKT-hopfield neural networks (KHN) for solving nonlinear optimization problems that contain inequality constraints. In this work, the development and the testing of the KHN algorithm was carried out. The KHN algorithm was used for solving two engineering design problems which were; 'optimization of the design of a pressure vessel' (P1) and the 'optimization of the design of a tension/compression spring' (P2). The computational performance of the KHN algorithm was then compared against the modified particle swarm optimization (PSO) algorithm of previous work on similar engineering problems. Comparative studies and analysis were then carried out based on the optimized results.


Related Articles

  • Perspectives on Stochastic Optimization Over Time. Tsitsiklis, John N. // INFORMS Journal on Computing;Winter2010, Vol. 22 Issue 1, p18 

    In this article the author discusses the integration of stochastic optimization research into artificial intelligence (AI) and operations research (OR). He elaborates the connection between reinforcement learning and approximate dynamic programming (ADP). He states that the heuristic evaluation...

  • The MiniZinc Challenge 2008-2013. Stuckey, Peter J.; Feydy, Thibaut; Schutt, Andreas; Tack, Guido; Fischer, Julien // AI Magazine;Summer2014, Vol. 35 Issue 2, p55 

    MiniZinc is a solver-agnostic modeling language for defining and solving combinatorial satisfaction and optimization problems. MiniZinc provides a solver-independent modeling language that is now supported by constraint-programming solvers, mixed integer programming solvers, SAT and SAT modulo...

  • Heuristics for efficient classification. Fraughnaugh, Kathryn; Ryan, Jennifer; Zullo, Holly; Cox Jr., Louis Anthony // Annals of Operations Research;1998, Vol. 78 Issue 1-4, p189 

    The classification problem is to determine the class of an object when it is costly to observe the values of its attributes. This type of problem arises in fault diagnosis, in the design of interactive expert systems, in reliability analysis of coherent systems, in discriminant analysis of test...

  • Merging AI and OR to Solve High-Dimensional Stochastic Optimization Problems Using Approximate Dynamic Programming. Powell, Warren B. // INFORMS Journal on Computing;Winter2010, Vol. 22 Issue 1, p2 

    We consider the problem of optimizing over time hundreds or thousands of discrete entities that may be characterized by relatively complex attributes, in the presence of different forms of uncertainty. Such problems arise in a range of operational settings such as transportation and logistics,...

  • Editorial: Special Section on Reinforcement Learning and Approximate Dynamic Programming. Xin Xu // Journal of Intelligent Learning Systems & Applications;May2010, Vol. 2 Issue 2, p55 

    The author presents his view on three research papers regarding the behavioral adaptability of algorithm for sequential decision making in the area in conjunction to approximate dynamic programming (ADP) and reinforcement learning (RL) for machines. It focuses on the potential application like...

  • Efficient methods for pseudo-tree based distributed best first search. Matsui, Toshihiro; Matsuo, Hiroshi // International MultiConference of Engineers & Computer Scientists;2007, p77 

    Distributed constraint optimization problem is an area of research in multi agent system. In recent years, a distributed constraint optimization algorithm, which performs best-first search in bottom up manner according to pseudo tree, was proposed. In this paper, we propose several efficient...

  • Structure Learning of Bayesian Network with Bee Triple-Population Evolution Strategies. Yungang Zhu; Dayou Liu; Haiyang Jia; Yuxiao Huang // International Journal of Advancements in Computing Technology;Nov2011, Vol. 3 Issue 10, p140 

    Recently evolutionary search has been investigated by lots of researchers as method for structure learning of Bayesian network. In order to avoid the problem of premature convergence and enhance the learning accuracy, in this paper we integrated evolutionary way of bees and proposed...

  • Reinforcement Learning: A Tutorial Survey and Recent Advances. Gosavi, Abhijit // INFORMS Journal on Computing;Spring2009, Vol. 21 Issue 2, p178 

    In the last few years, reinforcement learning (RL), also called adaptive (or approximate) dynamic programming, has emerged as a powerful tool for solving complex sequential decision-making problems in control theory. Although seminal research in this area was performed in the artificial...

  • Entropy Maximization through Dynamic Programming. Jha, P.; Chandravanshi, Anjali // Global Journal of Pure & Applied Mathematics;2012, Vol. 8 Issue 4, p415 

    Dynamic programming is an optimization technique based on Bellman's principle of optimality. In this paper we have maximized the Kapur's entropy by using dynamic programming under constraints ... .


Read the Article


Sorry, but this item is not currently available from your library.

Try another library?
Sign out of this library

Other Topics