Page - 11 - in Algorithms for Scheduling Problems
Image of the Page - 11 -
Text of the Page - 11 -
Algorithms 2018,11, 18
Algorithm1:Greedyinsertionheuristicalgorithmwithmulti-stagefilteringmechanism
1. Sortall jobs innon-increasingorderof theirpowerconsumptionrates
2. Initialization: Ik=bk+1−bk, forall 1≤k≤m
3.For i=1tondo
3.1. If layer1
C1. IfCondition1 issatisfied
Initial theperiod indexkk=argmink∈A(Ik≥ ti)
//Job i isprocessedwithinperiodkk.
C2.Else ifCondition2 issatisfied
//Job i isprocessedacrossperiodskandk+1.
3.2.Else if layer2
C3. IfCondition3 issatisfied
C3.1. If inequality (8) isnotsatisfied
//Job i isprocessedacrossperiodsk,k+1,andk+2.
C3.2.Else
Initial theperiod indexkk′=argmink′∈B(Ik′ ≥ ti)
//Job i isprocessedwithinperiodkk’.
C4.Else ifCondition4 issatisfied
C4.1. Ifdk+1=0
//Calculate cost1, cost3,and cost4andinsert job i into thepositionwithminimal insertioncost.
C4.2.Else ifdk+2=0anddk+1>0
//Calculate cost1, cost2, cost3, cost4,and cost5andinsert job i into thepositionwithminimal
insertioncost.
C5.Else ifCondition5 issatisfied
Initial theperiod indexkk′=argmink′∈B(Ik′ ≥ ti)
//Job i isprocessedwithinperiodkk’.
3.3.Else if layer3
C6. IfCondition6 issatisfied
//Similarly toCondition4, itneeds tocalculate the insertioncostof severalpossiblepositionsandinsert
job i into thepositionwithminimal insertioncost.
C7.Else ifCondition7 issatisfied
C7.1. Ifmaxk′′∈Γ {
Ik′′ }
> ti
Initial theperiod indexkk′′=argmink′′∈Γ (
Ik′′ ≥ ti )
//Job i isprocessedwithinperiodkk”.
C7.2.Else
//Job i traversesallnon-fullon-peakperiodsandinsert job i into thepositionwithminimal
insertioncost.
4.ComputationalResults
In thissection,areal-life instancefromamachinerymanufacturingcompanyinChinaisprovided
to further illustrate theMILPmodel and the proposed algorithm. Then, two sets of contrasting
experimentswithrandomlygenerated instancesareconducted,aimingtoshowthegoodperformance
of the algorithm.The algorithm is coded inMATLABR2015b and the experiments are runon an
Intel(R)Core(TM) i7-47903.60GHzprocessorwith16GBofmemoryunder theWindows7operating
system. For benchmarking, the greedy insertionheuristic algorithmproposedbyChe et al. [9] is
adoptedforourcontrast tests.
TheTOUelectricity tariffsusedforall instancesare those implementedinShanxiProvince,China,
asgiven inTable1. It canbeseenfromTable1 that theoff-peakperiod isbetweenanon-peakperiod
andamid-peakperiod,whichmeans that the actual electricitypricesmeets thefirst typeofTOU
electricity tariffs.Assumethat the timehorizonstartsat8:00a.m. of thefirstday.
11
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