Abstract
In the presented work foldable containers are used to transport wood products from forest sites to sawmills and further on to customers. In order to solve this overall planning problem a combined vehicle routing and container scheduling system is developed and formulated as a Mixed Integer Programming model. The solution process of the problem is state-dependent during execution and integrates empty container management and vehicle routing. Real life sized problem instances are solved with Tabu Search based solution methods that are both compared to each other and to results obtained with solver software. Extensive numerical studies were made to test the model and prove the good performance of the presented solution approaches. The initial solution of the metaheuristics is computed with a modified nearest neighbour method. Especially, the dynamic Tabu Search with Alternating Strategy, where neighbourhoods are changed dynamically, shows a good performance on required solution quality.
Special thanks go to Dr. Hamish Waterer for inspiring discussions on model formulations and motivating literature support as well as to Prof. David Ryan and Assoc. Prof. Matthias Ehrgott for discussions on that topic. Thanks also go to Marco Oberscheider for assistance in performing numerical tests and programming.