Constructing Runge–Kutta methods with the use of artificial neural networks

A methodology that can generate the optimal coefficients of a numerical method with the use of an artificial neural network is presented in this work. The network can be designed to produce a finite difference algorithm that solves a specific system of ordinary differential equations numerically. The case we are examining here concerns an explicit two-stage Runge–Kutta method for the numerical solution of the two-body problem. Following the implementation of the network, the latter is trained to obtain the optimal values for the coefficients of the Runge–Kutta method. The comparison of the new method to others that are well known in the literature proves its efficiency and demonstrates the capability of the network to provide efficient algorithms for specific problems.

[1]  Simon Haykin,et al.  Neural Networks: A Comprehensive Foundation , 1998 .

[2]  R. Tetzlaff,et al.  A learning algorithm for cellular neural networks (CNN) solving nonlinear partial differential equations , 1995, Proceedings of ISSE'95 - International Symposium on Signals, Systems and Electronics.

[3]  Luo Siwei,et al.  Numerical solution of differential equations by radial basis function neural networks , 2002, Proceedings of the 2002 International Joint Conference on Neural Networks. IJCNN'02 (Cat. No.02CH37290).

[4]  Andrew J. Meade,et al.  Solution of nonlinear ordinary differential equations by feedforward neural networks , 1994 .

[5]  J. Butcher.  Numerical methods for ordinary differential equations , 2003 .

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

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

[8]  Bin-Da Liu,et al.  A backpropagation algorithm with adaptive learning rate and momentum coefficient , 2002, Proceedings of the 2002 International Joint Conference on Neural Networks. IJCNN'02 (Cat. No.02CH37290).

[9]  N. Phan-Thien,et al.  Neural-network-based approximations for solving partial differential equations , 1994 .

[10]  George Cybenko,et al.  Approximation by superpositions of a sigmoidal function , 1989, Math. Control. Signals Syst..

[11]  J. Butcher.  Numerical Methods for Ordinary Differential Equations: Butcher/Numerical Methods , 2005 .

[12]  Euripidis Glavas,et al.  Solving differential equations with constructed neural networks , 2009, Neurocomputing.

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

[14]  Matematik,et al.  Numerical Methods for Ordinary Differential Equations: Butcher/Numerical Methods , 2005 .

[15]  Charalampos Tsitouras,et al.  Neural networks with multidimensional transfer functions , 2002, IEEE Trans. Neural Networks.

[16]  Kurt Hornik,et al.  Approximation capabilities of multilayer feedforward networks , 1991, Neural Networks.

[17]  Haralambos Sarimveis,et al.  Radial Basis Function Network Training Using a Nonsymmetric Partition of the Input Space and Particle Swarm Optimization , 2013, IEEE Transactions on Neural Networks and Learning Systems.

[18]  Satish S. Udpa,et al.  Finite-element neural networks for solving differential equations , 2005, IEEE Transactions on Neural Networks.