Gerenciamento da Malha Ferroviária de uma Usina Siderúrgica : Propostas para a Simulação de Evendos, Modelagem de Processos, e o Planejamento de Rotas Baseado na Metaheurística de Colônia de Formigas

• Ano de defesa: 2019
• Autor(a) principal: Ferreira, Vinicius dos Reis Alves
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Palavras-chave em Português: Otimização por colônia de formigas; Modelo térmico; Usina siderúrgica; Roteamento de veículos; Transporte de gusa
• Link de acesso: http://www.repositorio.ufc.br/handle/riufc/54936

Process optimization and cost reduction are key interrelated topics for the survival of any business. More and more resources are being invested in automation and continuous improvement of industrial processes focused on operational excellence. Based on this, the present work proposes the reduction of costs related to the losses of the pig iron transportation process in an integrated steel mill. In this type of steel mill, the pig iron produced in blast furnace is transported to steelmaking plant, where it will be transformed into steel. This transport is carried out on the most of steel mills by rail vehicles such as locomotives and torpedo ladle cars. The distance traveled and the time spent in this process have a direct impact on the loss of pig iron thermal energy to the environment, which implies higher costs for steel production, besides transportation costs. This work proposes a model based on the vehicle routing problem for the optimization of railway traffic management and pig iron transportation in a steel mill. A thermal model is also created to estimate the temperature of the metal and torpedo ladle car during transport between points where temperature measurement is not possible. To solve the problem, it was chose a metaheuristic based on ant colony optimization applied to vehicle routing and to reducing temperature variation of the torpedo ladle car, which seeks to reduce temperature losses and also the wear of refractory material that occurs due to these variations. The models were tested using a simulator developed and the results compared with operational data of a Brazilian steel mill whose main product is the manufacture of steel slabs. Data related to the performance and applicability of the algorithm in production environment are also discussed. The main results achieved by the algorithm include a 17% reduction in the total distance traveled, in addition to a 21 and 25% reduction in thermal losses for empty and charged torpedo ladle cars, respectively, which represented an approximate 40°C increase is the average temperature of the pig iron delivered to the steelmaking plant against the results obtained through the actions taken by experienced operators.