Tabu Search for Optimization of Military Supply Distribution

The dynamic and non-hierarchical nature of the military domain presents a challenge for traditional supply chain optimization. Flow networks and simulation techniques have been applied to the military distribution problem, but are unable to provide time-definite delivery to customers. Recently, optimization approaches have been independently applied towards strategic and operational levels of planning. However, decomposing military distribution into separate problems forces optimization techniques to utilize imprecise data. The size of the military distribution problem has prevented optimization techniques from providing end-to-end planning capabilities. This paper presents a Tabu Search algorithm for simultaneously solving strategic and operational levels of planning. The algorithm uses partial-order planning to separate the optimization process from the constraint verification process. The problem is reduced to a tractable computation by representing scenarios as two-tier systems and only permitting transshipments between different tiers. The results verify that the presented algorithm discovers higher quality solutions than simulation for simultaneously solving strategic and operational levels of planning.