The Instances of AGV and machine integrated scheduling problem in flexible job shop
The problem corresponding to the Instances set is the AGV and machine integrated scheduling problem in the flexible job shop, or the flexible job shop scheduling problem with AGV, or the flexible job shop scheduling problem considering transportation time.
The specific description is as follows :
n Jobs need to be processed on m machines, and the intermediate material transportation is completed by K AGVs ; each job contains n_j operations, and each operation can be processed by any machine from a given optional machine set. The processing time of the operation will vary with the different machine ; the transfer of jobs between machines is realized by a limited number of AGVs.For the transfer of any job, the time of AGV's action can be devided into pickup time and delivery time. The scheduling goal is to determine each operation processing machine, arrange the appropriate operation processing sequence and assign the job to the appropriate AGV for transportation, so as to minimize the total completion time.
The first data set is 10 small instances modified by Bilge and Ulusoy [ 1 ] proposed by Deroussi and Norre [ 2], which is denoted as fjsp1 : 10, located in the file fjspT.py.
The second data set uses 48 small examples proposed by Kumer et al. [ 70 ].These 48 small examples are divided into two groups according to the ratio of transportation time t to processing time p. The first group is t / p > 0.25, which is denoted as EX ( t / p > 0.25 ). The second group is, this group is realized by multiplying the processing time of the original example by 2 and dividing the transportation time by 2, located in file EX.py
The third data set is obtained by extending the FJSP example proposed by Fattahi [ 71 ], which is denoted as Fattahi _ Set, located in DataSet1/FJSP_Fttahi,Other data sets in the Data _ Set1 file can also be used.
In all the above instances, the number of AGVs is 2, and need to combine the file M4 _ to _ M18.py and Generator_FJSPT.py to load single instances.
Loading the above example uses the function Load_Instance_FJSPT() in Generator_FJSPT.py
In the benchmark instances, only the case of 2 AGVs is considered. In order to further verify the effectiveness of the algorithm under different AGV numbers, this paper also generates the following three random instances :
(1) 12 small-scale instances, where the number of jobs is J = { 10,15,20 }, the number of machines is M = { 4,6,8 }, the number of AGVs is R = { 2,3 }, and the small-scale instances are located in file small_scale ;
(2) 20 medium-sized instances, where the number of jobs is J = { 30,40,50 }, the number of machines is M = { 8,10,15 }, the number of AGVs is R = { 3,4,6 }, the medium-sized instances are located in file middle_scale ;
(3) 27 large-scale instances, where the number of jobs is J = { 60,80,100 }, the number of machines is M = { 15,20,25 }, the number of AGVs is R = { 4,6,8 }, and the large-scale instances are located in file large_scale.
Loading the above example uses the function Load_Instance_Data_Set1() in Generator_FJSPT.py