부산대학교 도서관

We consider a class of stochastic optimal control problems for discrete-time linear systems whose objective is the characterization of control policies that will steer the probability distribution of the terminal state of the system close to a desired Gaussian distribution. In our problem formulation, the closeness between the terminal state distribution and the desired (goal) distribution is measured in terms of the squared Wasser-stein distance which is associated with a corresponding terminal cost term. We recast the stochastic optimal control problem as a finite-dimensional nonlinear program whose performance index can be expressed as the difference of two convex functions. This representation of the performance index allows us to find local minimizers of the original nonlinear program via the socalled convex-concave procedure [1]. Finally, we present nontrivial numerical simulations to demonstrate the efficacy of the proposed technique by comparing it with sequential quadratic programming methods in terms of computation time.