Intelligent job shop scheduling via deep reinforcement learning over graphs
Job Shop Scheduling Problem (JSSP) is a well-known NP-hard combinatorial optimization problem (COP) with extensive applications in today’s manufacturing system. Due to its NP-hardness, approximation, heuristic, and meta-heuristic algorithms have been proposed in the past. These methods have some li...
Saved in:
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Thesis-Doctor of Philosophy |
| Language: | English |
| Published: |
Nanyang Technological University
2023
|
| Subjects: | |
| Online Access: | https://hdl.handle.net/10356/164926 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Nanyang Technological University |
| Language: | English |
| Summary: | Job Shop Scheduling Problem (JSSP) is a well-known NP-hard combinatorial optimization problem (COP) with extensive applications in today’s manufacturing
system. Due to its NP-hardness, approximation, heuristic, and meta-heuristic algorithms have been proposed in the past. These methods have some limitations,
among which two are well recognized. One is high computation overhead due to
the nature of computational inefficiency of the methods and the curse of dimensionality (the problem sizes). The other is that existing methods strongly depend
on human expert experience for algorithm design, which is less automatic. In addition, the manually designed components are also highly dependent on human
expertise, lacking a substantial level of exploration. |
|---|