(A) multiple pass heuristic scheduling algorithm for the resource constainted project scheduling problem

Abstract

In this study, an effective heuristic algorithm for the problems of scheduling activities under resource and precedence constraints has been developed. The proposed algorithm which starts with an initial feasible solution using Thesen[18]``s algorithm (i.e., one-pass type) attempts to reduce the project completion time by a search heuristic. The heuristic reschedules activities in the search region in order to find improvements utilizing flexibilities within a schedule. The search regions are determined iteratively in the search procedure. Though the characteristics of activity slacks (or floats), when the resources are limited, is very complicated, it is still valuable to consider activity slacks for a measurement of flexibility in a project schedule. It is possible to move an activity back and forth within its slack. The combination of backward and forward moving operations (right and left shifts) may result in a reduction of the project duration. It is noted, in the middle of rescheduling, feasibility is always maintained and finishing time of each activity does not exceed the current project completion time. A random problem generator and a random schedule generator are used for the evaluation of this algorithm. Computational results indicate that the multiple pass algorithm works better for the problems having relatively less number of precedence relationship.