Stabilization of discrete-time systems with stochastic parameters (Q1057827)

Relying on linear quadratic optimal control methods, the author proposes a sufficient and easy to check algebraic condition for a discrete-time stochastic system, described by the equation \(x_{k+1}=a_ kAx_ k+b_ kBu_ k,\) to be stochastic stabilizable in both senses of mean- square exponential and almost surely asymptotic stability, where the state variable \(x_ k\in R^ n\), the control variable \(u_ k\in R^ m\), A and B are constant matrices of appropriate dimensions, \(a_ k\) and \(b_ k\) are time-invariant random sequences, not depending on the initial condition \(x_ 0\) and such that \(E\{a^ 2_ k\}=E\{b^ 2_ k\}=1\) and \(E\{a_ kb_ k\}=E\{a_ 0b_ 0\}\) for all k. \(E\{\) \(\cdot \}\) denotes the expectation operator. A constructive method for stabilizing gains is also given.