Machine Learning prediction model for Dynamic Scheduling of Hybrid Flow-Shop based on Metaheuristic
Abstract
Optimizing scheduling and allocation strategies in dynamic production environments, notably in Hybrid Flow-Shops, presents significant challenges. This study focuses on resource assignment within dynamic contexts. It proposes an approach that use Genetic Algorithm (GA) to generate data and train machine Learning (ML) to predict near optimal allocations. Through experiments across various scenarios, the accuracy of prediction of different ML models for resource allocation is evaluated. Our findings highlight the potential of ML techniques to improve decision-making in dynamic and flexible manufacturing systems (FMS), contributing to efforts to enhance reactive scheduling strategies. Future work will assess the impact of these decisions on mean completion time, providing deeper insights into on-line scheduling efficiency.
Origin | Files produced by the author(s) |
---|