Implementation of the Metaheuristic Tabu Search in Route Selection for Mobility Analysis Support System

This thesis employs a reactive tabu search heuristic implemented in the Java programming language to solve a real world variation of the vehicle routing problem with the objective of providing quality routes to Mobility Analysis Support System (MASS). MASS is a stochastic simulation model used extensively by Air Mobility Command (AMC) to analyze strategic airlift capabilities and future procurement decisions. This dynamic real world problem of strategic and tactical airlift possesses a number of side constraints such as vehicle capacities, route length and time windows in a sizeable network with multiple depots and a large fleet of heterogeneous vehicles. Finding optimal solutions to this problem is currently not practical. Currently, MASS requires all possible routes used in its simulation to be manually selected. As a result, the route selection process is a tedious and time consuming process that relies on experience and past performance of the model to obtain quality routes for the mobility system.