| المستخلص: |
The scheduling of the multi-project consists of two or more projects, and the project is defined as a collection of activities that consume resources or a set of events. Therefore, the scheduling of a project is the assignment of a start and finish time for each activity so that the precedence and resource requirements are satisfied. This becomes a difficult problem in multi-project environments with limited resource. Three choices for scheduling the projects are usually considered. The first is when the project is very large-scale. Therefore, the project should be divided into several large to medium-scale project so that each project can be easily scheduled. The second choice is when there are several small projects, which have nearly converged completion time, then the treating them separately may lead to lose in efforts, time and to inefficient management and improper control. Therefore, the solution is scheduling these projects in multi-project management environments. The third one is the dealing of several medium scaled projects, which can be managed separately with easy scheduling and high flexibility. In this thesis, the second choice is considered by taking five different construction projects in terms of their resources quantity and of the almost equal number of activities from a real construction company. However, their completion times are un-similar. Several approaches of scheduling these projects as a single multi-project will be presented and discussed. A genetic algorithm is used to develop a program that assists to solve the optimization scheduling process. Two different alternatives of scheduling will be proposed. The first alternative is merging the separate projects in one project by considered seven different merging configurations. The second alternative is re-planning activities of these projects in a single multi-project. For selecting the optimum trade-off between these scheduling alternatives, a multi-criterion is properly used. At first place, the thesis provides a general approach to scheduling several projects in the multi-project environments, and shows how to choose the best alternative of scheduling. Considering the case study, a general finding is although one optimal scheduling alternative (optimal merging) is found; however, this may not be the best scheduling or best merging for another case study. Consequently, managers have to determine their own multi-criterion for the case at his or her hands. The found optimal scheduling alternative (optimal merging) offers the best resource improvement coefficient RIC, maximum resource rate RR, and minimum resource maximal use MR, and the risk considerations. Furthermore, the alternative of seven merged configurations is compared to the alternative of re-planning the activities of all projects in a single multi-project. It is found that, besides the need of manager experiences and skills to reduction the number of total activities for an alternative of re-planning the activities of all projects in a single multi-project; the results are either similar or less optimal.
|
