Solving a multiple objective linear program using simulated annealing

Multiple Objective Linear Programs (MOLP) are often used in practice for managerial decision making. There are no well-accepted Operations Research techniques that can provide optimum solutions for MOLP. This paper reviews some of the popular MOLP techniques currently reported in the literature indicating their advantages and disadvantages. In addition, we investigate the use of simulated annealing for solving a constrained maximization problem with two objectives. The design of the algorithm is presented and experimental experiences are reported. Finally, the requirements of resources for different candidate solutions are discussed in the light of defining the term optimality in multiobjective optimization.