Ingeniare. Revista chilena de ingeniería
versión On-line ISSN 0718-3305
PEREZ RAVE, Jorge Iván; JARAMILLO ALVAREZ, Gloria Patricia; PARRA MESA, Carlos Mario y MORENO VELASQUEZ, Luis Fernando. SYSTEMIC TETRAHEURISTIC FOR THE TSP. Ingeniare. Rev. chil. ing. [online]. 2010, vol.18, n.2, pp. 187-202. ISSN 0718-3305. doi: 10.4067/S0718-33052010000200006.
This paper presents a novel method to solve instances of the TSP. This method is comparable in effectiveness and efficiency with "nearest neighbour", "cheapest insertion", "two-way exchange improvement" and "branch and bound". The first section provides a literature review of the combinatorial optimization, the second provides a reference frame, the third the methodology used and the fourth contains, inter alia, system thinking, AxB factorial design and management tool CAP-DO. The fourth section presents the design "new" method called systemic tetraheuristic, followed by ANOVA and Duncan ranges for each factor: method and number of cities; this section concludes with an analysis of the proportion of "failures" of the proposed algorithm with increasing complexity of the TSP. As a result of this study a method is presented for solving the TSP. This method consists of: 1- "nearest neighbour", 2- "short-term sacrifice" 3- "LIFO transfer" and a support heuristic "right search 4P4". For the design of the "sacrifice short-term" procedure, it was necessary to analyze the "Nearest Neighbour" heuristic from a systems perspective. This heuristic is related to the "quick solutions that fail" archetype, which can be applied to day-to-day. The proposed method has proved to be more efficient than others, in what concerns quality of the solution and the computation time, especially when the complexity of the TSP increases.
Palabras clave : Combinatorial optimization; tetraheuristics; systems thinking; travelling salesman problem.