Dynamic programming models with goal objectives

From MaRDI portal
Publication:3323719





The concepts of goal programming (GP) are applied to some multiple objective problems which have the characteristics of dynamic programming (DP). Following the concepts of GP, a target goal is set for each objective. In GP, instead of trying to optimize the objective function directly, deviations between goals and what can be achieved subject to a given set of constraints are minimized. In the solution process, goals are achieved according to their importance. Deviations from the goals are ranked according to 'pre-emptive' priority factors. The problem in question has N stages and J target goals for each stage. The objective function of the GP problem depends on the variables representing the deviations from the respective target goals and on their pre-emptive priority levels. The problem has both goal and system constraints. A dynamic programming formulation of the problem in question is applicable if the problem can be decomposed into N equivalent subproblems, for each priority level, each containing only one decision variable and M state variables. The generalization of the formulation given in the article to include multiple decision variables at each stage, is straightforward.











This page was built for publication: Dynamic programming models with goal objectives

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3323719)