A Cultural Approach to Job Sequencing Optimization

摘要

Our intent is to improve the performance of a paint shop sequencing operation through the application of evolutionary techniques. The environment in which this application operates is complex, involving the optimization of multiple hard and soft constraints. In the past, deterministic solutions have been applied to non-deterministic problems. This environment is non-deterministic making it well suited to an evolutionary approach. In this paper, we present an approach to global optimization which will reduce overall sequence disruption, minimize work in progress and minimize paint changes through blocking. Consideration must be given to the anomalies which cause disruptions to the production sequence. Each of the goals are not complementary and sacrifices must be made in each goal in order to gain an optimal outcome. Optimization of the build queue to satisfy the constraints can easily be approximated to problems that are evaluated to np-complete. Our aim is to show how our evolutionary approach can be applied to this problem.