Reproduction of the iterative three-component heuristic for the team orienteering problem with time windows

This thesis reproduces the Iterative 3-Component Heuristic (I3CH) of Hu and Lim (2014). The I3CH tries to solve the Team Orienteering Problem with Time Windows (TOPTW). The aim of the TOPTW is to maximize the total profit score obtained by visiting a set of locations, while each location can only be visited in its time window, with a limited number of vehicles. The I3CH consists three components to solve the TOPTW. A local search and simulated anneal-ing procedure are used to search routes in the solution space. The route recombination procedure uses mathematical programming to select a combination of routes found in the local search and simulated annealing procedure such that the profit score is maximized. Our reproduction of the I3CH performs 0.11% better than the I3CH of Hu and Lim (2014), but our heuristic is slower than theirs. The profit scores of 87 out of 304 instances are improved. We found 16 solutions with higher profit scores than the best known solutions available in 2012 and 10 solutions with higher profit scores than the latest best known solutions (Gunawan, Lau, and Lu, 2015). Moreover, this thesis extends the TOPTW by taking into account that certain locations cannot be visited by a route. The heuristic is changed in multiple ways, which leads to lower computation times.