Aplicação de heurísticas para o sequenciamento dinâmico de atividades em máquina única para a minimização do custo de multas por atraso

Abstract

Operational research has been applied to production planning and control in several areas, whether services or consumer goods. Through optimization of key elements such as logistics, costs, better distribution of resources, sequencing of activities or processes and among others, we have a better performance in the market. Because of those variables, this work aimed to study a scenario in which the industries are subject to the application of fines for delayed tasks not delivered in the determined period. In a productive environment composed of a single machine, tasks were assigned to be fulfilled in certain periods, each with a fine related to the delay of its delivery. For the implementation, an improvement heuristic was applied in the C programming language. In 99.25% of instances there was a reduction in the cost of late fines when compared to the initial sequencing. In total 800 experiments were performed, with 20 to 1000 activities. Cost savings from applying the improvement amounted to 52.5%.