Nonseparable Dynamic Programming in Discrete Time Systems

Recent research has revealed that a class of nonseparable dynamic optimization problems can be embedded into a separable multiobjective optimization setting which can be effectively solved by multiobjective dynamic programming. The purpose of this paper is to establish a general framework for solving nonseparable discrete-time dynamic optimization problems using multiobjective dynamic programing. Especially, the general conditions to characterize the optimal solution of a nonseparable dynamic problem from among the set of noninferior solutions of the corresponding maltiobjective dynamic programming problem are established.