International Journal of Intelligent Systems and Applications (IJISA)
IJISA Vol. 3, No. 5, 8 Aug. 2011
Cover page and Table of Contents: PDF (size: 294KB)
Computing Method and Hardware Circuit Implementation of Neural Network on Finite Element Analysis
Full Text (PDF, 294KB), PP.41-47
Views: 0 Downloads: 0
Index Terms
Finite element method, Hopfield neural network, analogue circuit, simulation, operational amplifier, digital controlled potentiometer
Abstract
The finite element analysis in theory of elasticity is corresponded to the quadratic programming with equality constraint, which can be further transformed into the unconstrained optimization. In the paper, the neural network of finite element solving was obtained on the basis of Hopfield neural network that was reformed. And the no error solving of finite element neural net computation was realized in theory. And a design method to construct an artificial neuron by using electronic devices such as operational amplifier, digital controlled potentiometer and so on was presented. A programmable hardware neural network of finite element can be build up by using analog switches to interconnect inputs/outputs of hardware neurons. The weights, biases and connection in the hardware neural network of finite element can be adjusted automatically by microprocessor according to the results of train to controlling system, This programmable hardware neural network of finite element has some more adaptability for different systems. In addition, the authors present the computer simulation and analogue circuit experiment to verify this method. The results are revealed that: 1) The results of improved Hopfield neural network are reliable and accuracy; 2) The improved Hopfield neural network model has an advantage on circuit realization and the computing time, which is unrelated with complexity of the structure, is constant. It is practical significance for the research and calculation.
Cite This Paper
Likun Cui, Wei Wang, Zhuo Li, "Computing Method and Hardware Circuit Implementation of Neural Network on Finite Element Analysis", International Journal of Intelligent Systems and Applications(IJISA), vol.3, no.5, pp.41-47, 2011. DOI:10.5815/ijisa.2011.05.06
Reference
[1]N. G. Xie, T. T. Yu, W. D. Shao, “Network parallel algorithm in optimization computation of engineering structure," Journal of East China University of Metallargy, vol.18, no. 2, pp. 117-120, 2001 (in Chinese).
[2]W. L. Zhang. "Studying status and development of computing mechanics", Journal of Anhui institute of architecture, vol.9, no. 2, pp. 7-12, 2001 (in Chinese).
[3]X. Chen, D. H. Sun, H. Z .Huang, "Finite element system for structural analysis and neural network", Hoisting and conveying machinery, no. 6, pp. 6-8, 1999 (in Chinese).
[4]X. C. Wang ," Finite element method”, Beijing: Tsinghua University Press,2003 (in Chinese).
[5]D. H. Sun, Q. Hu, H. Xu ," Real-time Neurocomputing Model Used in Elastic Mechanics", China mechanical engineering,vol.8, no. 2, pp. 35-38, 1997 (in Chinese).
[6]H. B. Li, Neurocomputing based structural finite element analysis, Dalian: Dalian university of technology, 2003 (in Chinese).
[7]L. K. Cui, W. Wang, Z. Li, "Application of Hopfield neural network in finite element solving", Computer Engineering and Applications,vol.46, no. 16, pp. 46-47+49, 2010 (in Chinese).
[8]H. Z. Huang , H. B. Li, "Research on neurocomputing method on finite element analysis", Journal of Mechanical Strength,vol.25, no. 3, pp. 298-301, 2003 (in Chinese).
[9]E. Z. Liang, E. W. Liang, Circuit design and simulation application base on Protel 99 SE ,Beijing: Tsinghua University Press,2000 (in Chinese).
[10]Datasheet:“<0.4Ω CMOS 1.8V to 5.5V, SPST Switches ADG801/ADG802”. Analog Devices