Authors:
Oliver Polo-Mejía
1
;
Christian Artigues
2
and
Pierre Lopez
2
Affiliations:
1
CEA, DEN, DEC, SETC, St. Paul lez Durance, France, LAAS-CNRS, Université de Toulouse, CNRS, Toulouse and France
;
2
LAAS-CNRS, Université de Toulouse, CNRS, Toulouse and France
Keyword(s):
Project Scheduling, Multi-skill, Partial Preemption, Nuclear Research Facility.
Related
Ontology
Subjects/Areas/Topics:
Applications
;
Artificial Intelligence
;
Enterprise Information Systems
;
Information Systems Analysis and Specification
;
Knowledge Discovery and Information Retrieval
;
Knowledge-Based Systems
;
Methodologies and Technologies
;
Operational Research
;
Optimization
;
Pattern Recognition
;
Project Management
;
Scheduling
;
Software Engineering
;
Symbolic Systems
Abstract:
In this article we consider a new scheduling problem known as the Multi-Skill Project Scheduling Problem with Partial Preemption. The main characteristic of this problem is the way we handle the resources release during the preemption periods: only a subset of resources are released. Since this problem is NP-hard, we propose a greedy algorithm based on priority rules, modeling the subproblem of technicians allocation as a Minimum-Cost Maximum-Flow problem. In order to improve the performance of the greedy algorithm, we propose a randomized tree-based local search algorithm. Computational tests are carried out and analyzed.