Co-evolutionary approach to the manpower planning problem

Abstract

Abstract. This dissertation will investigate a realistically sized manpower planning problem. To perform the analysis, an appropriate model will be pre- sented to solve a speci c type of manpower problem. It allows for recruitment, dismissal, training of workers to di erent types, and multi{skilled workers. The training is enforced in the model through time delay. The model also includes consistency, demand and bound constraints that the solutions must satisfy. The solution of the model will then be investigated using a novel approach, the competitive co{evolutionary algorithm. The algorithm generates solutions in the same way a genetic algorithm does, and implements a predator{prey procedure to ensure satisfaction of the constraints. A numerical example will then be solved to illustrate the usefulness of the co{evolutionary algorithm when applied to this type of manpower problem.