Composite random sampling algorithm for scheduling concurrent projects

Yong-yi SHOU
DOI: https://doi.org/10.3785/j.issn.1008-973X.2006.02.036
2006-01-01
Abstract:In order to overcome the shortcoming of traditional priority rule based heuristics, a new composite random sampling method based on serial schedule generation scheme was proposed to increase the effectiveness of resource allocation and shorten the overall project durations. The method extends the serial schedule generation scheme used in single project scheduling to multi-project scheduling problems. Using the regret-based biased random sampling technique, the method combines various priority rules to schedule a set of simultaneous projects repetitively, so as to select the best multi-project schedule. Systematic experimental tests show that the proposed composite random sampling method improves the resource allocation among project activities and significantly shortens the overall project durations.
What problem does this paper attempt to address?