Exact controllability of linear stochastic impulsive systems

From MaRDI portal





The authors consider systems with both continuous and discrete control inputs and use backward stochastic differential equations as the main tool for the analysis. A central step of the paper is the transformation of the original forward stochastic impulsive system into an associated backward stochastic impulsive system. This is achieved by decomposing the control input under suitable rank assumptions on the diffusion control matrix. The solution of the backward system is then expressed recursively by means of an adjoint matrix-valued stochastic differential equation and conditional expectation arguments.\N\NThe main results provide rank-type controllability criteria in terms of Gramian matrices which combine the effects of the continuous dynamics, the impulse operator, and the discrete control input. In particular, the paper gives sufficient full-rank conditions for exact controllability and a necessary rank condition involving the combined continuous and discrete controllability matrices. The approach is also consistent with known deterministic controllability criteria for impulsive systems when stochastic perturbations are absent.\N\NThe authors further extend the method to stochastic impulsive systems with input delays. In this delayed case, modified controllability Gramians are introduced, and analogous sufficient and necessary rank conditions are obtained. Two numerical examples, based on the Euler--Maruyama method and Monte Carlo computation of the Gramians, illustrate the applicability of the proposed criteria and show that exact controllability may still hold even when control is applied only on selected impulse intervals.











This page was built for publication: Exact controllability of linear stochastic impulsive systems

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q7260473)