COMPARATIVA DE ALGORITMOS PARA LA CONSTRUCCIÓN DE GRAFOS PERT DE PROYECTOS GRANDES

Abstract

The well known PERT method for project planning uses AOA (Activity On Arc) graphs. Preparing these graphs requires the inclusion of dummy activities. Ideally, the smaller the number of dummy activities, the smaller the time of computation taken on following calculations, and the greater the understandability. However, the preparation of an optimal graph (one with the lowest number of dummy activities) is a high computational complexity problem (NP class). In order to work with large projects, there are diverse algorithms that reduce the computation time by allowing the inclusion of more dummy activities than necessary. Including recent algorithms, we make a comparison focused on performance: the number of dummy activities included in relation to the computational cost.