Identification of parameters of interval nonlinear models of static systems using multidimensional optimization

The article proposes an approach to parametric identification of interval nonlinear models of static systems based on the standard problem of minimizing the root mean square deviation between the values ​​of the modeled characteristics of the static object and the values ​​belonging to the experimental intervals. As a result of expanding the parameter space of nonlinear models by introducing additional coefficients to match the predicted and experimental values into the objective function, a multidimensional optimization problem with a nonlinear multiextremal objective function is obtained. The paper examines the characteristics of the objective function and the convergence of its optimization.