The use of neural network approach for solving of tasks of estimation of parameters and state variables

As for identification automatic control systems the systems with constraints on the measurements are particularly important. This problem is known as a joint estimation of parameters and statе variables. The approaches to this problem are analyzed in this paper. The shortcomings of the classical approach to system identification are revealed. The neural network approach for solving the problem of joint estimation of parameters and statе variables is proposed. The schemes of artificial neural network implementation are considered.