.png){kind=icon}
Return
A simheuristic-based algorithm for the stochastic long-term maintenance scheduling problem
.png){kind=icon}
OA
AI
DOI：10.1111/itor.70021.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
This work addresses the problem of assigning preventive maintenance jobs in a 52-week planning horizon. Given a set of machines that need preventive maintenance, a set of maintenance jobs in these machines, a set of work teams, and a planning horizon, the problem consists of assigning each job to a work team in a given instant of the planning horizon, aiming to minimize the cost with work teams and the cost of performing the unscheduled jobs using outsourced teams. We propose an iterated local search (ILS)-based algorithm specialized for this problem. Using real instances, the ILS algorithm achieved the best results in 81% of the instances, outperforming literature algorithms. However, these algorithms only treat the deterministic version of the problem and do not consider the uncertainty in the job duration that may occur in an industry environment. Not considering this aspect can produce an inefficient schedule with many unscheduled jobs. So, this work also proposes a simheuristic-based algorithm (SIM-ILS) capable of capturing this issue. We tested it in three scenarios, which differ in the level of uncertainty regarding the job duration, and compared their results with those provided by the stochastically evaluated ILS solutions. SIM-ILS found the best solution in 61% of the tests. Therefore, the SIM-ILS can be used to support decision-making in different industrial environments, from environments with low variability in job duration to those with high variability.
Keywords:
long-term preventive maintenance scheduling
iterated local search
variable neighborhood descent
metaheuristics
dispatching rules
simheuristic
Journal
IF:
2.9
Papers:
1.8K
Citations:
3.7K
Organization
Cited Papers
No cited papers available