Solution of Differential - Algebraic Equation Systems Using Neural Networks

In this paper, the solution of differential algebraic equations (DAES) is obtained using neural networks. The goal is to obtain accurate solution with reduced calculus effort by comparing the solution of the DAES obtained from Pade approximation and neural network method, the solution of DAE system is computed by feedforward neural network (FFNN). The solution of the neural network approach to this problem is very close to the exact solution. An illustrative numerical method is presented for the proposed method.