# Dataset
This is the dataset for the paper with title “A hierarchical multi-action deep reinforcement learning method for dynamic distributed job-shop scheduling problem with job arrivals” submitted to IEEE TRANSACTIONS ON AUTOMATION SCIENCE AND ENGINEERING. 

The raw data for the three scenarios from the automotive engine manufacturing company in the case study are stored in “Scenario1”, “Scenario2” and “Scenario3”, respectively. For example, the data for Scenario 1 is stored in “Scenario1” file.

For any file “Scenario[Scenario No.]”, three Excel files and two PNG files are included.
data_arrival_[number of initial jobs]_[number of arrival jobs].xlsx
data_ops_[number of initial jobs]_ [number of arrival jobs] .xlsx
data_pts_[number of initial jobs]_ [number of arrival jobs] .xlsx
Scenario[Scenario No.]_GanttChart_shopfloor_1.png
Scenario[Scenario No.]_GanttChart_shopfloor_2.png


In “data_arrival_[number of initial jobs]_[number of arrival jobs].xlsx” file, the data are stored as 
|  Column 1   |  Column 2  |  Column 3  |
|  ----  | ----  |  ----  |
| Job index  | Part Type |         arrival time |


In “data_ops_[number of initial jobs]_ [number of arrival jobs] .xlsx” file, the data are stored as 
Column 1	Column 2 										Column 3										...
Job index	the machine that the first operation of Job [Job index] is processed on 		the machine that the second operation of Job [Job index] is processed on		...	


In "data_pts_[number of initial jobs]_ [number of arrival jobs] .xlsx" file, the data are stored as
Column 1	Column 2 								Column 3								...
Job index	the processing time of the first operation of Job [Job index]		the processing time of the second operation of Job [Job index]		...


File "Scenario[Scenario No.]_GanttChart_shopfloor_[Factory No.].png" presents the Gantt Chart of the scheduling scheme generated by the proposed MDRL method in factory [Factory No.].