Neural network based semi-empirical models for dynamical systems described by differential-algebraic equations

We analyzed the problem of mathematical modeling and computer simulation of nonlinear controlled dynamical systems usually described by differential-algebraic equations. The problem is proposed to be solved in the framework of the semi-empirical approach combining theoretical knowledge for the plant with training tools of artificial neural network field. The results are presented for a semi-empirical model that simulate the reentry hypersonic vehicle and confirm the efficiency of this approach.