Modelos de otimização para o sequenciamento de cirurgias no médioprazo

• Ano de defesa: 2015
• Autor(a) principal: Virgínia Giani Casagrande
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Minas Gerais; UFMG
• Palavras-chave em Português: Métodos heurísticos; Horizonte de tempo de médio prazo; Sequenciamento de cirurgias; Cirurgia Modelos matemáticos; Engenharia de produção
• Link de acesso: http://hdl.handle.net/1843/BUBD-AAHHYK

The surgery schedule problem can be seen in the health area, in whichthe operating theatre management for the patients treatment in a hospitalvaries according to different specialties and involves physical, humanand informational resources. In this context, the surgery schedule process must ensure the surgical realization procedures according to medical constraints, the time window of each surgery and the available demand in a hospital. This work addresses the surgery scheduling through integer programming mathematical models and the respective heuristic application methods with these models. The problem modeling is carried out for the medium-term time horizon and aims to optimize a performance evaluation criteria, maximizing occupancy or minimizing tardiness, considering the resources competition in a hospital operating theatre. The problem evaluated in this study is called NP-hard, whereas finding a valid solution can be considered as NP-complete. Heuristics based on the integer programming model are developed on two levels. On the first level, a pre-allocation of surgery is assigned to the days of the planning horizon. On the second level each day is settled as an independent sub-problem. Computational experiments are presented in a real case based instances to evaluate the modelling attempts and to compare the mathematical model with the heuristics approaches. Each approach executes the relaxed model for one of the objective functions, carrying on the data from linear relaxation to get the results. Among the four proposed approaches, the first two ones stand out by showing promising results and several opportunities for conducting future work.