This paperaddresses thebatchscheduling problem ofunrelated parallel machines attempting tominimize thetotal weighted tardiness. Identical orsimilar jobsare typically processed inbatches todecrease setupand/or processing times. Localdispatching rules suchastheearliest weighted duedate, theshortest weighted processing time, and theearliest weighted duedatewithaprocess utilization spread aretailored tothebatchscheduling requirements. Basedonthe features ofbatchscheduling, a two-level batchscheduling frameworkissuggested. Existing heuristics, whichshow excellent performance intermsoftotal weighted tardiness for thesingle machine scheduling, suchasthemodified earliest due dateruleandthemodified costovertimerule, areextended for theproblem. The simulated annealing algorithm as a meta-heuristic isalsopresented toobtainnearoptimal solutions. Theproposed heuristics arecomparedthrough computational experiments withdatafromthedicing process ofacompound semiconductor manufacturing facility IndexTerms-Scheduling, parallel machine, heuristics, tardiness
[1]
Chris N. Potts,et al.
A decomposition algorithm for the single machine total tardiness problem
,
1982,
Oper. Res. Lett..
[2]
Christos Koulamas,et al.
The Total Tardiness Problem: Review and Extensions
,
1994,
Oper. Res..
[3]
C.-H. Kuo,et al.
Evaluating scheduling heuristics for non-identical parallel processors
,
1997
.
[4]
Bahram Alidaee,et al.
Scheduling parallel machines to minimize total weighted and unweighted tardiness
,
1997,
Comput. Oper. Res..
[5]
Hamilton Emmons,et al.
Scheduling families of jobs with setup times
,
1997
.
[6]
T.C.E. Cheng,et al.
A state-of-the-art review of parallel-machine scheduling research
,
1990
.