Home // INTELLI 2025, The Fourteenth International Conference on Intelligent Systems and Applications // View article
Authors:
Erik De Kuyffer
Toon De Pessemier
Wout Joseph
Luc Martens
Keywords: VRPTW; VRPy; OR Tools; ACO; Job Shop Scheduling; Pareto.
Abstract:
The high fuel prices and the important costs associated with windmill downtime during maintenance urge the need to minimize travel time and scheduling of jobs in a short time period. Since landing in windmills at sea is difficult and depends on meteorological parameters, the constraint of maintenance windows is added when searching for the optimal route. To minimize the distance traveled, the Vehicle Routing Problem with Time Windows (VRPTW) is solved, using three different methods. The VRPTW is applied to two separate databases, namely various sets of windmills to be maintained and several numbers of customers to be serviced. Applications with 8 to 175 windmills, divided over 3 farms have shown that the VRPTW solved by using three different methods resulted in a similar relative gain in travel distance, compared to a randomly chosen route. The main difference between the methods studied is the amount of calculation time needed, which varies from 1 second to 6 minutes for the different methods. To demonstrate the general applicability, the same three methods were executed on a set of service tasks performed at 8 to 40 customers of a window decoration company, distributed throughout Belgium, resulting in similar outcomes. In a second part of the paper, the Job Shop Scheduling Problem (JSSP) is solved to minimize the total maintenance span of offshore windmills as an additional objective function. This led to a relative gain of up to 62% in maintenance time, compared to the total maximum maintenance span for an application of 40 windmills. Finally, both objectives, minimal distance and minimal maintenance time span, are combined, resulting in a set of non-dominated maintenance sequences that can be used by the planner.
Pages: 30 to 38
Copyright: Copyright (c) IARIA, 2025
Publication date: March 9, 2025
Published in: conference
ISSN: 2308-4065
ISBN: 978-1-68558-236-4
Location: Lisbon, Portugal
Dates: from March 9, 2025 to March 13, 2025