Optimal Superexponential Stabilization of Solutions of Linear Stochastic Differential Equations

Abstract

We consider the problem of superexponential stabilization of a scalar linear controlled stochastic process. The underlying stochastic differential equation contains both additive and multiplicative disturbance terms. To achieve stabilization, an infinite-time control problem is solved. The cost is assumed to be quadratic and having a superexponentially increasing time-weighting function. We study the convergence of the optimal process to a zero state in the mean-square sense and almost surely.