Page - 119 - in Algorithms for Scheduling Problems
Image of the Page - 119 -
Text of the Page - 119 -
Algorithms 2018,11, 50
Theenumeration isorganizedbyshifting theoperationsof the lastorder to the left insuchaway
that they trade theirplaceswithpreviouscompetingoperationsofprecedingorders (seeFigure3).
Assoonastheoperationreaches its leftmost feasiblepositionandthecurrentcombinationcorresponds
to a linear programming problem with no solution we roll back to the previous combination,
stopshiftingoperationsof the lastorderandswitch to theoperationsof thepreviousorder.After the
enumeration for orders on the current resource is over, we proceed the same iterationswith the
next resource. Theformalpresentationof thedesignedbranch-and-boundalgorithmisdescribedby
Procedure1.
Procedure1: branch-and-bound
1.FindtheinitialsolutionoftheLPproblem (1)ā(3)forthecombinationckā{0,1}, k=1,. . . ,K
thatcorresponds to thecasewhenallordersare inarow(ļ¬rstoperationof the followingorderstarts
onlyafter the lastoperationofprecedingorder iscompleted).
2.Remember thesolutionandkeepthevalueofobjective functionΦas temporarilybest result
Opt=Φ.
3.Select the lastorder f= |F|.
4.Select theresource r= |R| that isusedbylastoperation i= |J|of theprecedencegraphGof
themanufacturingprocess.
5.Thebranch< i,r, f> isnowformulated.
6.Condition:Does theselectedresource rhavemorethanoneoperation inthemanufacturing
procedure thatallocates it?
6.1. Ifyes then
Beginevaluatingbranch< i,r, f>
Shift theoperation ionecompetingpositionto the left
Find the solution of theLPproblem (1)ā(3) for current combination and calculate the
objective functionΦ
Condition: is thesolutionfeasible?
If feasible then
Condition: Is theobjective functionvalueΦbetter thantemporarilybest resultOpt?
Ifyes then
savecurrentsolutionas thenewbest resultOpt=Φ.
Endofcondition
Switch to the preceding operation i = iā1 of currently selected order f for the
currentlyselectedresource r.
Goto thep. 5andstartevaluatingthenewbranch
Ifnot feasible then
Stopevaluatingbranch< i,r, f>
Switch to theprecedingresource r= rā1
Goto thep. 5andstartevaluatingthenewbranch
Endofcondition: is thesolutionfeasible?
Endofcondition: p. 6
7.Switchto theprecedingorder f= fā1
8.Repeatpp. 4ā7untilnomorebranches< i,r, f>areavailable
9.Repeatpp. 3ā8untilnomore feasibleshiftsarepossible foralloperations inallbranches.
Iterating combinations for one order on one resource can be considered as evaluating one
branchof thealgorithm[10]. Finding the leftmostposition for anoperation is similar todetecting
a bound. Switching between branches and stopping on bounds is formalized with a general
branch-and-boundprocedure.
119
back to the
book Algorithms for Scheduling Problems"
Algorithms for Scheduling Problems
- Title
- Algorithms for Scheduling Problems
- Authors
- Frank Werner
- Larysa Burtseva
- Yuri Sotskov
- Editor
- MDPI
- Location
- Basel
- Date
- 2018
- Language
- English
- License
- CC BY 4.0
- ISBN
- 978-3-03897-120-7
- Size
- 17.0 x 24.4 cm
- Pages
- 212
- Keywords
- Scheduling Problems in Logistics, Transport, Timetabling, Sports, Healthcare, Engineering, Energy Management
- Categories
- Informatik
- Technik