Abordagem híbrida para o problema de escalonamento de enfermeiros

• Autor(a) principal: Garcia, Renan Leon
• Data de Publicação: 2016
• Tipo de documento: Dissertação
• Idioma: por
• Título da fonte: Repositório Institucional da Universidade Estadual de Maringá (RI-UEM)
• Texto Completo: http://repositorio.uem.br:8080/jspui/handle/1/2489
• topic: Problema de escalonamento de enfermeiros; Atribuição de serviços; Escalonamento; Enfermeiros; Meta-heurística; Algoritmos heurísticos; Método de busca local; Problema de Atribuição; Heurística; Brasil.; Nurse scheduling problem; Assignment problem; Heuristic; Meta-heuristic; Brazil.; Ciências Exatas e da Terra; Ciência da Computação

This dissertation addresses a Nurses Scheduling Problem (NSP), which consists in constructing a working schedule for nurses in a hospital taken account their preferences and the constraints imposed by the problem. For each nurse is assigned a sequence of shifts covering the entire schedule period. The NSP is an optimization problem classified as Non Polinomial Hard (NP-hard). It has fostered the development of several models and algorithms, heuristics and meta-heuristics, due to the great difficulty of getting a good solution. This work proposes a hybrid approach to NSP involving a variation of Variable Neighborhood Search (VNS) meta-heuristic, combining heuristic method to explore the neighborhood, with exact methods to generate neighbor solution. As local search process three methods were used to implement the neighborhood structures, called Cut and Recombination Process (CRP), k-swap and Cover Fit. To validate the proposal is used a database with 24 real-world instances of the problem. The results obtained showed that the proposed approach performs consistently well for small/medium size instances, where the results obtained are compared with results presented by other works that used the same database.