1 / 23

Hybrid Algorithm for mean T

Hybrid Algorithm for mean T. Baker p.49. Hybrid Algorithm : Dominance matrix to decompose problem in to sub-problems. Branch and Bound. Original problem for DP n! = C(n,1) + C(n,2) + ………….+ C(n,n) 30~35 jobs → out of computation time. N!. N 1 !.

rozene
Download Presentation

Hybrid Algorithm for mean T

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Hybrid Algorithm for mean T Baker p.49 Hybrid Algorithm: Dominance matrix to decompose problem in to sub-problems. Branch and Bound Original problem for DP n! = C(n,1) + C(n,2) + ………….+ C(n,n) 30~35 jobs → out of computation time N! N1! N2!

  2. Decomposition Ex. Divide 50 persons into 4 subgroups by according to their height. Mutual exclusive S Decompose S1,S2,………,Sn and For any , put i in Then the objective function

  3. Hybrid Algorithm Notation: SPT EDD

  4. Bj tj Hybrid Algorithm Property 3.2: (i before j) Property 3.3: (j before i)

  5. Hybrid Matrix

  6. Steps: Baker p.51

  7. Ex. (See Baker p.52)

  8. Final table 2-5-1-4-6-7-XXX-10 389

  9. By DP 2-5-1-4-6-7-9-3-8-10

  10. HW. Use Hybrid to solve

  11. Midterm

  12. Single Machine Heuristics Sule p.14 • Earliest Due Date (EDD) Rule • Cost Over Time (COVERT) Rule • Shortest Processing Time (SPT) Rule • Largest Penalty per Unit Length (LPUL) Rule • Largest Weight (WT) and LPUL Rule • Shortest Processing Time and LPUL Rule • Shortest Weighted Processing Time (SWPT) Rule • Critical Ratio (CR) Rule

  13. COVERT Notations: TT - Total processing times of all jobs RT - Remaining total processing times of the set of unscheduled jobs ST - Starting time for the next scheduled job,0 for the first job CF - Coefficient PR - Priority

  14. PR=0 PR=1 Di TT Pi ST+Pi ST COVERT Rules Step l: calculate PR for all jobs RT Step 2: calculate CF for job i, yet to be scheduled. Step 3: Job with the Max CF is scheduled next. Repeat step1

  15. Ex. Iteration 1. Step l : Caculate PR TT=93 RT=93

  16. Ex. Step 2: Step3: job 3 is scheduled. Iteration 2. . . . .

  17. Backward & Forward Method • Backward&ForwardMethod Initial scheduled Improve:Swapping to jump out the local optimal • Use any one of the Lower Bound Heuristic method • Not the optimum

  18. BF Method 5 jobs X X X X X 5 4 3 2 1 4: 5 – 1 3: 5 – 2 , 4 – 1 2: 5 – 3 , 4 – 2 , 3 – 1 1: 5 – 4 , 4 – 3 , 3 – 2 , 2 –1 Lag =

  19. Ex. (1) Backward Phase: Step1. TT = 37+27+1+28 = 93 Calculate the penalty for each job if it is to be complete at T=93. Job1: (93-49)*1=44 Job2: (93-36)*5=285 Job3: (93-1)*1=92 Job4: (93-37)*5=280 • job 1 has the least cost, so it is scheduled in the last position

  20. EX. Step2. T= 93-37 = 56 Calculate the penalty for each job if it is to be complete at T=56. Job2: (56-36)*5=100 Job3: (56-1)*1=55 Job4: (56-37)*5=95 • job 3 has the least cost, so it is scheduled in position 3

  21. EX. Step3. T= 56-1 = 55 Calculate the penalty for each job if it is to be complete at T=55. Job2: (55-36)*5=95 Job4: (55-37)*5=90 • job 4 has the least cost, so it is scheduled in position 2 Step4. The final sequence 2-4-3-1 The total penalty for the sequence 2-4-3-1 is (18*5)+(55*1)+(44*1)=189

  22. Ex. (2) Forward Phase: Initial sequence : 2-4-3-1 penalty = 189 Lag = 3 : 1-4-3-2 penalty = 490 – no change Lag = 2 : 3-4-2-1 penalty = 144 – change New sequence : 3-4-2-1 penalty = 144 Lag = 3 : 1-4-2-3 penalty = 644 – no change Lag = 2 : 2-4-3-1 penalty = 189 – no change Lag = 2 : 3-1-2-4 penalty > 144 – no change Lag = 1 : 4-3-2-1 penalty = 172 – no change Lag = 1 : 3-2-4-1 penalty = 139 – change New sequence : 3-2-4-1 . . . We can get the final sequence is 3-2-4-1

  23. HW. • Use Forward + Covert to solve the following problem : (or you want to compare your solution with BF)

More Related