Bi-level dynamic scheduling architecture based on service unit digital twin agents
| Autor: | Haojie Chen, Deng Tingming, Jian Zhang, Guofu Ding, Shengfeng Qin, Jiang Haifan |
|---|---|
| Rok vydání: | 2021 |
| Předmět: |
0209 industrial biotechnology
Service (systems architecture) business.industry Computer science Emerging technologies G400 Distributed computing Big data Scheduling (production processes) Cloud computing 02 engineering and technology Dynamic priority scheduling G600 Industrial and Manufacturing Engineering Unit (housing) 020901 industrial engineering & automation Hardware and Architecture Control and Systems Engineering 0202 electrical engineering electronic engineering information engineering 020201 artificial intelligence & image processing Architecture business W200 Software |
| Zdroj: | Journal of Manufacturing Systems. 60:59-79 |
| ISSN: | 0278-6125 |
| DOI: | 10.1016/j.jmsy.2021.05.007 |
| Popis: | Pure reactive scheduling is one of the core technologies to solve the complex dynamic disturbance factors in real-time. The emergence of CPS, digital twin, cloud computing, big data and other new technologies based on the industrial Internet enables information acquisition and pure reactive scheduling more practical to some extent. However, how to build a new architecture to solve the problems which traditional dynamic scheduling methods cannot solve becomes a new research challenge. Therefore, this paper designs a new bi-level distributed dynamic workshop scheduling architecture, which is based on the workshop digital twin scheduling agent and multiple service unit digital twin scheduling agents.\ud \ud Within this architecture, scheduling a physical workshop is decomposed to the whole workshop scheduling in the first level and its service unit scheduling in the second level. On the first level, the whole workshop scheduling is executed by its virtual workshop coordination (scheduling) agent embedded with the workshop digital twin consisting of multi-service unit digital twins. On the second level, each service unit scheduling coordinated by the first level scheduling is executed in a distributed way by the corresponding service unit scheduling agent associated with its service unit digital twin. The benefits of the new architecture include (1) if a dynamic scheduling only requires a single service unit scheduling, it will then be performed in the corresponding service unit scheduling without involving other service units, which will make the scheduling locally, simply and robustly. (2) when a dynamic scheduling requires changes in multiple service units in a coordinated way, the first level scheduling will be executed and then coordinate the second level service unit scheduling accordingly. This divide-and-then-conquer strategy will make the scheduling easier and practical.\ud \ud The proposed architecture has been tested to illustrate its feasibility and practicality. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|
Full Text from ScienceDirect