Development of a suboptimal method for the synthesis of control algorithms

Abstract

The computation of the trace of the Fisher information matrix in this work is carried out in terms of the sensitivity functions of the output variables. It is essential that the computation of the corresponding sensitivity functions is performed in a feedback loop, which leads to the application of a dynamic programming method that differs from the standard scheme. The difference lies in the need to solve a system of linear algebraic equations in the synthesis of the control sequence. The use of such a control synthesis scheme in the work is justified, and necessary and sufficient conditions for the existence and uniqueness of the solution of the specified system are obtained. The distribution of the energy of identification signals between more and less important parameters of the model is varied by changing the weighting factors. It is proposed to impose an additional constraint on the trace of Fisher's information matrix on the control sequence synthesized on the basis of the principle of deterministic equivalence.