Dimensionamento e sequenciamento de lotes de produção na indústria de bens de consumo de higiene pessoal
AGRADECIMENTOS À minha esposa Luciana, pelo carinho, pelo apoio e pela compreensão em todos esses anos. Aos meus pais e irmãos, por todas as conquistas da minha vida. À professora Débora, pela orientação, pela dedicação e pelo constante apoio emocional. Aos professores Reinaldo e Franklina, pelas valiosas contribuições a esse trabalho concedidas no Exame de Qualificação. Aos colegas, professores e funcionários do programa de pós-graduação, pela ótima convivência nesse período. ABSTRACT This
... . ABSTRACT This work adresses the integrated lot sizing and scheduling problem. The process of lot sizing and scheduling consists of determining how much to produce of each product and the scheduling of these lots in each machine in order to meet the demand under existing restrictions and operational capabilities. The case studied in this work describes the scenario of a big company in the industry of consumer goods for personal hygiene, a very competitive market in which the good service level for customers and the cost management show up in the competition for the clients. In this company, there is a complex operational environment, composed of distinct parallel machines with limited production capacity and sequence dependente setup times. There is also a limited finished goods storage capacity, a characteristic not found in the existing literature. Backordering is allowed but it is extremely undesirable. This problem is NP-hard and it has been treated by heuristics in the literature. In this work, we developed nine different solving strategies based on the relax-and-fix heuristics. The aim of this approach is not only to solve a complex real problem but also assess whether the form of partitioning and the sequence of solving the subproblems influences the performance of the relax-and-fix heuristics. The computational experiments were conducted on ramdomly generated instances and real problems. The results showed the good performance and the robustness of the proposed approach. Alternative techniques were applied in the strategy with the best results in the previous tests to enhance its performance.