Publication: Comparing the performances of priority dispatching rules and q-learning through production simulation
Loading...
Date
2023-07-01
Authors
Nurul Husna Binti Mohd Daman Huri
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
In practical production scheduling, objectives are typically derived from desired production performance factors withing a manufacturing system. However, customer-oriented production priority management emphasizes the analysis of operational production management policies that are challenging to quantify. By conducting a production simulation, this study aims to evaluate the effectiveness and efficiency or priority dispatching rules and Q-learning in improving production performance when demonstrating the adaptive rule selection strategy in meet desired objectives in dynamic manufacturing environment. The findings from this contribute to the understanding of how different approaches to production prioritization impact overall manufacturing system performance.
The application of simple dispatching rules is essential for assuring their precision and effectiveness. Data envelopment analysis is used to find a set of rules that conform to the scheduling requirements to handle this issue. Additionally, depending on various situations produced from dispatching rule, Q-learning and customer-oriented production priority rules, the analytic hierarchy method is used to prioritise work.
Performance measurements are produced by illustrating a production system where parts are organized into batches and given precise lot sizes to evaluate the proposed approach for different scenarios when sorting for received order and sequence of order. The dispatching rules created utilizing the suggested technique to other dispatching rules in terms of their performance measured by key parameters that give insight into the manufacturer’s fulfilment like station utilization, work-in-progress, average lateness, flow time ratio and global objective. The outcome shows that the suggested strategy performed satisfactorily in increasing production effectiveness and achieving the targeted goals and can identify the most effective rules that align with the specific production objectives and criteria. Basically, this thesis gives valuable insights into the comparison between priority dispatching rules and Q-learning in the context of production simulation