Improving Scheduling In Labeling Production Enviroment With Particle Swarm Optimization

Abstract

This study proposes a non-linear integer model applying onto the environmental labeling industry. The model is considering penalty of earliness or lateness delivery, with portion of profit loss due to outsourcing and the penalty cost impacted from the unbalanced of production. This thesis also proposes an optimization technique by applying the Particle Swarm Optimization with a repair encoding scheme which regenerate the infeasible solution from the simulation of Particle Swarm Optimization. The performance of parameters of Particle Swarm Optimization are analyzed. The study also analyzes the priority of matching the inventory with the orders. A dataset from labeling manufacturer are experimented with computational coding Particle Swarm Optimization and repairing encoding scheme. The result is shown to be valid and applicable with the labeling manufacturing environment.