Séminaire en format hybride au GERAD local 4488 ou Zoom

We address the maintenance planning for rotor blades on onshore wind turbines at a tactical level, motivated by the case of a German onshore wind turbine manufacturer. The goal is to select maintenance teams from external service providers and assign maintenance tasks for a maintenance season while minimizing the total costs. The operative scheduling of the maintenance tasks and routing of the maintenance teams are anticipated to determine the capacity needs.

We consider team-specific work time regulations, locations, qualifications, cost rates, means of transport, task-specific time windows, and weather conditions. However, planning such a maintenance season is very challenging. For this reason, we formulate a mixed-integer linear problem for routing and scheduling the heterogeneous teams to perform maintenance tasks at different locations with minimum total costs. We propose a branch-and-price algorithm approach based on a Dantzig-Wolfe decomposition to solve it. The pricing subproblems are solved using a labeling algorithm.

The approach is implemented using the SCIP framework. We present the underlying problem decomposition, implementation, and first numerical results using a state-of-the-art commercial solver as a benchmark.