Las presentaciones del seminario giraron en torno a problemas relevantes relacionados con el área de transporte, presentando estrategias para su solución.
Durante la tarde del miércoles 04, se realizó el Seminario de Investigación, instancia precursora en su categoría, ya que el evento fue abierto a la comunidad.
La instancia contó con el apoyo del director de la Escuela de Ingeniería en Transporte, Juan Sánchez y de la Jefa de Docencia, María Isabel Vega, para facilitar la participación de los alumnos y fue organizada por la académica Lorena Bearzotti y por el profesor Javier Maturana.
Al respecto de los objetivos de esta actividad, la profesora Bearzotti señala que, “la realización del seminario de investigación persigue dos objetivos claros, el primero es favorecer la creación de redes de investigación tanto nacionales como internacionales, y el segundo es potenciar la investigación en los estudiantes, para los cuales poder participar en este seminario significa una primera aproximación al mundo de la investigación escuchando a sus profesores en otros roles y también a profesores e investigadores de otras instituciones”.
La presencia de los profesores participantes y sus investigaciones se describen a continuación:
Douglas Smith (Universidad de Missouri, Saint Louis, Estados Unidos)
Título: Simulation modeling for collaborative planning of airport infrastructure and flight operations.
Abstract: With historical background on the forces which shaped the development of St. Louis Lambert International Airport, we discuss the challenge of right-sizing facilities and getting optimal use from airport assets as they approach their capacity limits. We then present a simulation model designed to foster collaboration among major airport stakeholders as they strive to balance airside and groundside activity, reduce delays and plan for future activity.
Javier Maturana (ITRA)
Título: Solving the Bi-Objective Insular Traveling Salesman Problem
Abstract: A new formulation for the traveling salesman problem (TSP) is presented. The subtour elimination constraint is strengthened by a lifting technique and valid inequalities are incorporated evaluating if decrease the search space between the linear relaxation and the convex envelope. Computational results reveal that the proposed formulation present better solutions than the three most effective formulations available in the literature. The instances have taken from the TSP library (involving between 17 and 443 cities) and the models were working within an hour of CPU time by Cplex.
Leopoldo López (Universidad de Talca)
Título: Evidence based production planning in Hospital Clinical Service
Abstract: The current production planning process in any Hospital Clinical Service is done with aggregate data. The planning does not use local information because this is very scarce, inaccurate or null, which may be useless for an adequate annual programming.
To use local information doctors, need to standardize the different pathologies by using an international notation like CIE-10. This work studied six month of dermatology patients in a Chilean Hospital. A doctor with CIE-10 experience rewrote the diagnostics, then the information was analyze.
The contribution was that it provided the Hospital with the epidemiological information necessary for a good programming and planning of a Dermatology Service. The study was presented to the local authorities who raised the Clinical Service Budget in five times.
Giovanni Campuzano (Universidad del Bío Bío)
Título: A New Formulation for the Traveling Salesman Problem
Abstract: A new formulation for the traveling salesman problem (TSP) is presented. The subtour elimination constraint is strengthened by a lifting technique and valid inequalities are incorporated evaluating if decrease the search space between the lineal relaxation and the convex envelope. Computational results reveal that the proposed formulation present better solutions than the three most effective formulations available in the literature. The instances have taken from the TSP library (involving between 17 and 443 cities) and the models was working within an hour of CPU time by Cplex.
Armando Guarnaschelli (ITRA)
Título: Network Design with Toll Optimization and Hazmat Transportation: A bi-level integrated approach
Abstract: In this talk, we introduce a bi-level integrated approach to simultaneously solve the Toll Optimization Problem, the Network Design Problem, and the Hazardous Material Transportation problem. Current research trends only provide partial modeling integrations. On the one hand, the network operator should consider every factor involved in new road investments, such as the flow of ordinary and hazmat vehicles and their toll prices. On the other hand, vehicles can choose among alternative routes based not only on travel time but also on costs due to toll charging and traffic congestion.
A mixed-integer non-linear bi-level problem is proposed to model this integration. In that problem, the network operator (leader) maximizes its benefit, defined as the toll income minus the costs due to the construction of new roads and the economic evaluation of risk exposure of people to hazmat transportation.
The vehicles (follower) minimize their travel costs including time, toll charging, and traffic congestion. The mixed-integer non-linear bi-level problem is approximated by a mixed-integer linear problem, and a binary search-based procedure is proposed to approach near-optimal toll levels.