Job Shop Scheduling with Setup Times and Time Windows


File format

The format is based on the t2-ps and t2-pss instances available in the Neil Simonetti's Traveling Salesman Problem Web Page.

Each bencnmark is composed by 25 instances with the following file names: dataps15MaxNN.txt or datapss15MaxNN.txt for the BTS25-TW15s; dataps2MaxNN.txt or datapss2MaxNN.txt for the BTS25-TW20s  (NN is the number of the instance).  For each instance (file):
  1. the first entry is the number of problems, always 1 for these files;
  2. the first entry of the second line is the number of jobs, call it j (the next sequence of "9999" is not used);
  3. the next j lines indicate how many steps are in each job, in these cases each entry is the number of machines m;
  4. the next j x m lines indicate the individual steps for each job (four entries for each line):
    1. machine number is first;
    2. processing time second;
    3. min separation time with the next activity, third;
    4. max separation time with the next activity, fourth;
    5. min and max separation times take the conventional value 0 for the last activity of each job and they are not used;
  5. the next (j x m + 1) squared entries are the set-up times.


Some of the best results obtained by the ISP procedure are available for the above  BTS25-TW15  and the BTS25-TW20 instances. The used alpha and beta parameters can be deduced from the names of the files. Each file represents a run of the ISP algorithm on the given benchmarck (BTS25-TW15 or BTS25-TW20), each file is organized over 6 columns. The first column represents the name of the instance, the second the initial makespan (infinity capacity solution), the third the solution's makspan, fourth and fifth columns are not used, the last and sixth column is the cpu time.