Allocation of maritime assets in surveillance missions

This thesis presents a way to optimize the allocation of surveillance units in an anti piracy mission. Because of the many similarities with search theory we have used the yet available research in this field to solve this problem. We assume that the total area to be surveyed consists of a finite set of non-overlapping subareas to which we can assign our surveillance units. We assume that the time during which we have to allocate our units consists of a finite set of periods with equal length. In these periods each of our surveillance units can be assigned to exactly one area. We introduce a measure indicates the risk (of pirate attacks) in a certain area. With respect to this value we try to optimize the allocation of our units, taking transit times between areas into account. In this thesis we use CPLEX to solve a mixed-integer linear formulation. We also consider two metaheuristics, namely cross-entropy and simulated annealing. The methods were tested on a variety of problems. CPLEX was able to solve most of these problems, except the larger ones. Simulated annealing finds near optimal solutions and is able to find a solution for all cases. Cross-entropy finds near optimal solutions for the small cases, but when increasing the number of periods it is unable to still find good solutions.