An artificial neural network method for solving boundary value problems: with arbitrary irregular boundary conditions

An artificial neural network (ANN) method was developed for solving boundary value problems (BVPs) on an arbitrary irregular domain in such a manner that all Dirichlet and/or Neuman boundary conditions (BCs) are automatically satisfied. Exact satisfaction of BCs is not available with traditional numerical solution techniques such as the finite element method (FEM). The ANN is trained by reducing error in the given differential equation (DE) at certain points within the domain. Selection of these points is significantly simpler than the often difficult definition of meshes for the FEM. The approximate solution is continuous and differentiable, and can be evaluated at any location in the domain independent of the set of points used for training. The continuous solution eliminates interpolation required of discrete solutions produced by the FEM. Reducing error in the DE at a particular location in the domain does not necessarily imply improvement in the approximate solution there. A theorem was developed, proving that the solution will improve whenever error in the DE is reduced at all locations in the domain during training. The actual training of ANNs reasonably approximates the assumptions required by the proof. This dissertation offers a significant contribution to the field by developing a method for solving BVPs where all BCs are automatically satisfied. It had already been established in the literature that such automatic BC satisfaction is beneficial when solving problems on rectangular domains, but this dissertation presents the first method applying the technique to irregular domain shapes. This was accomplished by developing an innovative length factor. Length factors ensure BC satisfaction extrapolate the values at Dirichlet boundaries into the domain, providing a solid starting point for ANN training to begin. The resulting method has been successful at solving even nonlinear and nonhomogenous BVPs to accuracy sufficient for typical engineering applications.

[1]  Dimitris G. Papageorgiou,et al.  Merlin-3.0. A multidimensional optimization environment , 1998 .

[2]  Kevin Stanley McFall,et al.  Artificial Neural Network Method for Solution of Boundary Value Problems With Exact Satisfaction of Arbitrary Boundary Conditions , 2009, IEEE Transactions on Neural Networks.

[3]  Nam Mai-Duy,et al.  Numerical solution of differential equations using multiquadric radial basis function networks , 2001, Neural Networks.

[4]  Richard S. Sutton,et al.  Landmark learning: An illustration of associative search , 1981, Biological Cybernetics.

[5]  I. B. Turksen,et al.  Fuzzy logic: review of recent concerns , 1997, 1997 IEEE International Conference on Systems, Man, and Cybernetics. Computational Cybernetics and Simulation.

[6]  J. Faires,et al.  Numerical Methods , 2002 .

[7]  David H. Sharp,et al.  Neural nets and artificial intelligence , 1989 .

[8]  D. Parisi,et al.  Solving differential equations with unsupervised neural networks , 2003 .

[9]  The domain of convergence for the iterative solution of nonlinear second order boundary value problems , 1986 .

[10]  Andrew J. Meade,et al.  The numerical solution of linear ordinary differential equations by feedforward neural networks , 1994 .

[11]  R. Agarwal.  Computational methods for discrete boundary value problems, II , 1992 .

[12]  H. Kardestuncer,et al.  Finite element handbook , 1987 .

[13]  G. Smith,et al.  Numerical Solution of Partial Differential Equations: Finite Difference Methods , 1978 .

[14]  Mohamed S. Kamel,et al.  Decision fusion in neural network ensembles , 2001, IJCNN'01. International Joint Conference on Neural Networks. Proceedings (Cat. No.01CH37222).

[15]  Bir Bhanu,et al.  Adaptive object detection from multisensor data , 1996, 1996 IEEE/SICE/RSJ International Conference on Multisensor Fusion and Integration for Intelligent Systems (Cat. No.96TH8242).

[16]  Mona E. Zaghloul,et al.  VLSI implementation of locally connected neural network for solving partial differential equations , 1996 .

[17]  Shonali Krishnaswamy,et al.  Mining data streams: a review , 2005, SGMD.

[18]  A. C. Tsoi.  Multilayer perceptron trained using radial basis functions , 1989 .

[19]  R. Agarwal,et al.  Monotone convergence of iterative methods for right focal point boundary value problems , 1988 .

[20]  Shiro Usui,et al.  Evolution and adaptation of neural networks , 2003, Proceedings of the International Joint Conference on Neural Networks, 2003..

[21]  Christian Goerick,et al.  Fast learning for problem classes using knowledge based network initialization , 2000, Proceedings of the IEEE-INNS-ENNS International Joint Conference on Neural Networks. IJCNN 2000. Neural Computing: New Challenges and Perspectives for the New Millennium.

[22]  Chuen-Tsai Sun,et al.  Neuro-fuzzy modeling and control , 1995, Proc. IEEE.

[23]  Tommy W. S. Chow,et al.  Feedforward networks training speed enhancement by optimal initialization of the synaptic coefficients , 2001, IEEE Trans. Neural Networks.

[24]  Siwei Luo,et al.  Numerical solution of elliptic partial differential equation using radial basis function neural networks , 2003, Neural Networks.

[25]  Kurt Hornik,et al.  Multilayer feedforward networks are universal approximators , 1989, Neural Networks.

[26]  James Demmel,et al.  Multigrid equation solvers for large-scale nonlinear finite element simulations , 1999 .

[27]  Lutz Prechelt,et al.  A quantitative study of experimental evaluations of neural network learning algorithms: Current research practice , 1996, Neural Networks.

[28]  Shu-Hsien Liao,et al.  Expert system methodologies and applications - a decade review from 1995 to 2004 , 2005, Expert Syst. Appl..

[29]  A. M. Turing,et al.  Computing Machinery and Intelligence , 1950, The Philosophy of Artificial Intelligence.

[30]  Magali R. G. Meireles,et al.  A comprehensive review for industrial applicability of artificial neural networks , 2003, IEEE Trans. Ind. Electron..

[31]  C. Lee Giles,et al.  Nonlinear dynamics of artificial neural systems , 1987 .

[32]  Exact matching of boundary conditions and incorporation of semiquantitative solution characteristics in initial approximations to boundary value problems , 1977 .

[33]  W. J. Gordon.  Blending-Function Methods of Bivariate and Multivariate Interpolation and Approximation , 1971 .

[34]  Martin A. Riedmiller,et al.  A direct adaptive method for faster backpropagation learning: the RPROP algorithm , 1993, IEEE International Conference on Neural Networks.

[35]  Juan Julián Merelo Guervós,et al.  Resource Review: A Web-Based Tour of Genetic Programming , 2004, Genetic Programming and Evolvable Machines.

[37]  T. Jankowski.  Multistep methods for nonlinear boundary-value problems with parameters , 1990 .

[38]  W. J. Gordon,et al.  Transfinite element methods: Blending-function interpolation over arbitrary curved element domains , 1973 .

[39]  B. Zhang,et al.  A neural-net approach to supervised learning of pole balancing , 1989, Proceedings. IEEE International Symposium on Intelligent Control 1989.

[40]  T. Leephakpreeda.  Novel determination of differential-equation solutions: universal approximation method , 2002 .

[41]  Dimitrios I. Fotiadis,et al.  Artificial neural networks for solving ordinary and partial differential equations , 1997, IEEE Trans. Neural Networks.

[42]  J. Z. Zhu,et al.  The finite element method , 1977 .

[43]  W. J. Gordon,et al.  Construction of curvilinear co-ordinate systems and applications to mesh generation , 1973 .

[44]  Konrad Reif,et al.  Multilayer neural networks for solving a class of partial differential equations , 2000, Neural Networks.

[45]  Dimitris G. Papageorgiou,et al.  Neural-network methods for boundary value problems with irregular boundaries , 2000, IEEE Trans. Neural Networks Learn. Syst..

[46]  S. Mukherjee,et al.  Boundary element techniques: Theory and applications in engineering , 1984 .