A chance-constrained stochastic inventory problem under imperfect information of state

This paper deals with a chance-constrained stochastic production-planning problem under hypotheses of imperfect information of inventory variables. Assuming a linear-Gaussian nature to the inventory balance system, the mean and covariance, statistical variables, are estimated from the Kalman filter equations. As a consequence, an approach — originally developed for stochastic problems under perfect information of state [8] — is adapted to provide sub-optimal solution for this problem. A simple example illustrates the basic ideas presented.