Two-phase heuristics for scheduling of multiple projects

Abstract

This paper deals with the problem of scheduling of multiple projects sharing a common pool of resources. New two-phase heuristics are proposed and compared with the existing single and two-phase heuristics. These two-phase heuristics are based on a two-stage prioritisation process of activities for resource allocation in which, at any decision point, the projects are first prioritised as per project selection rule and eligible activities in the projects are then prioritised as per activity selection rule. The two-phase heuristics are categorized into look-ahead and non-look-ahead type based on the project selection rules used. Performance of the heuristics is evaluated in two stages for two performance measures – minimising mean project delay and minimising increase over critical project duration. The results show that the some look-ahead heuristics produces better schedules when used with appropriate minimum/maximum criterion. The results also show that the heuristics which are superior in minimising mean project delay generally perform poorly in minimising increase over critical project duration.