English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Automatic industrial scheduling, aiming at optimizing the sequence of jobsover limited resources, is widely needed in manufacturing industries. However,existing scheduling systems heavily rely on heuristic algorithms, which eithergenerate ineffective solutions or compute inefficiently when job scaleincreases. Thus, it is of great importance to develop new large-scalealgorithms that are not only efficient and effective, but also capable ofsatisfying complex constraints in practice. In this paper, we propose a BilevelDeep reinforcement learning Scheduler, \textit{BDS}, in which the higher levelis responsible for exploring an initial global sequence, whereas the lowerlevel is aiming at exploitation for partial sequence refinements, and the twolevels are connected by a sliding-window sampling mechanism. In theimplementation, a Double Deep Q Network (DDQN) is used in the upper level andGraph Pointer Network (GPN) lies within the lower level. After the theoreticalguarantee for the convergence of BDS, we evaluate it in an industrial automaticwarehouse scenario, with job number up to $5000$ in each production line. It isshown that our proposed BDS significantly outperforms two most used heuristics,three strong deep networks, and another bilevel baseline approach. Inparticular, compared with the most used greedy-based heuristic algorithm inreal world which takes nearly an hour, our BDS can decrease the makespan by27.5\%, 28.6\% and 22.1\% for 3 largest datasets respectively, withcomputational time less than 200 seconds.

Bilevel Learning Model Towards Industrial Scheduling