Butenko, Sergiy2006-08-162017-04-072006-08-162017-04-072005-052006-08-16http://hdl.handle.net/1969.1/3735In this thesis we use greedy randomize adaptive search procedure (GRASP) to solve the traveling salesman problem (TSP). Starting with nearest neighbor method to construct the initial TSP tour, we apply the 2-opt and the path-relinking method for the initial tour improvement. To increase 2-opt search speed, fixed-radius near neighbor search and don0t − look bit techniques are introduced. For the same reason a new efficient data structure, the reverse array, is proposed to represent the TSP tour. Computational results show that GRASP gives fairly good solutions in a short time.en-USGreedy Randomized Adaptive Search Procedure (GRASP)Traveling Salesman Problem (TSP)Book