TY - JOUR
T1 - Load balancing in project assignment
AU - Liang, Zhirong
AU - Li, Yanzhi
AU - Lim, Andrew
AU - Guo, Songshan
PY - 2010/12
Y1 - 2010/12
N2 - In this paper, we study a project assignment problem. Specifically, a set of projects, each of which needs to be finished over a project development cycle, are to be assigned to a group of identical engineers over a discrete planning horizon. The workload of the projects is different and fluctuates over their development cycles. In each period, engineers have a maximum allowed workload. The objective of the problem is to assign the projects to engineers with the objective of balancing the total workload among the engineers; the load balance is measured by the difference between the maximum and the minimum total workload. Such a problem is new to the literature. After proving the problem is strongly NP-hard, we propose a two-stage heuristic approach to solve it. Extensive numerical experiments show that the proposed approach can achieve optimal or nearly optimal solutions for all test cases; such performance is much better than what can be obtained from an IP model solved with ILOG CPLEX. An analysis of the algorithm has also been provided to explain how the superior performance has been achieved. © 2010 Elsevier Ltd. All rights reserved.
AB - In this paper, we study a project assignment problem. Specifically, a set of projects, each of which needs to be finished over a project development cycle, are to be assigned to a group of identical engineers over a discrete planning horizon. The workload of the projects is different and fluctuates over their development cycles. In each period, engineers have a maximum allowed workload. The objective of the problem is to assign the projects to engineers with the objective of balancing the total workload among the engineers; the load balance is measured by the difference between the maximum and the minimum total workload. Such a problem is new to the literature. After proving the problem is strongly NP-hard, we propose a two-stage heuristic approach to solve it. Extensive numerical experiments show that the proposed approach can achieve optimal or nearly optimal solutions for all test cases; such performance is much better than what can be obtained from an IP model solved with ILOG CPLEX. An analysis of the algorithm has also been provided to explain how the superior performance has been achieved. © 2010 Elsevier Ltd. All rights reserved.
KW - Assignment
KW - Generalized assignment problem
KW - Load/workload balancing
KW - Project assignment
UR - http://www.scopus.com/inward/record.url?scp=77955276026&partnerID=8YFLogxK
UR - https://www.scopus.com/record/pubmetrics.uri?eid=2-s2.0-77955276026&origin=recordpage
U2 - 10.1016/j.cor.2010.03.016
DO - 10.1016/j.cor.2010.03.016
M3 - RGC 21 - Publication in refereed journal
SN - 0305-0548
VL - 37
SP - 2248
EP - 2256
JO - Computers and Operations Research
JF - Computers and Operations Research
IS - 12
ER -