Sunteți pe pagina 1din 14

Product based cells

Models
Minimize number of machines With cycle times Cost Unidirectional flow Multiple products

1 1 2 3 4 5 6 1 1 1 1

4 1

6 1

1 1

1 1

1 1

1 1

1 1

1 1 1 1 1 1

X3 = X5= 1 with Z = 2

1 1 2 3 4 5 6 4 3 3 2

4 4

6 4

1 3

1 2

3 5

4 2

2 1

3 5 2 1 3 5

CT = 6 X3 = X4 = X5 = 1. Y41 = Y52= Y53 = Y34 =Y35 = Y56 = 1

There is a limit of 6 minutes on each machine. The costs of the six machines are 40, 35, 50, 60, 80 and 30.

The objective function is to Minimize 40X1 + 35X2 +50X3 + 60X4 + 80X5 + 30X6. The optimum solution is given by X1 = X2 =X6 = 1 with Z = 105. Operation 6 goes to machine 1, operations 2, 3, and 5 go to machine 2 and operations 1 and 4 go to machine 6. The cycle times are 4, 6 and 6 respectively.

to explore if the cycle time can be reduced further for the given number of machines

Minimize u.

X
i 1

Y
i 1
n

ij

1j
X i ui

p
j 1

ij

The optimum solution is given by X3 = X4 =X5 = 1 with Z = 5. Operations 4 and 5 go to machine 3, operations 1 and 3 go to machine 4 and operations 2 and 6 go to machine 5. The cycle times are 5, 5 and 5 respectively and the maximum cycle time is 5.

Yij Xi . Xi, Yij = 0,1

Minimizing cost subject to cycle time and number of machines

The optimum solution is given by X1 = X2 =X6 = 1 with Z = 105. Operation 6 goes to machine 1, operations 2, 3, and 5 go to machine 2 and operations 1 and 4 go to machine 6. The cycle times are 4, 5 and 6 respectively.

Unidirectional flow
Let Zk = 1 if position k is chosen Let Xik = 1 if machine I goes to position k Let Yjk = 1 if operation j is carried out in position k m The constraints are m

Y
k 1

jk

1j

X
i 1

ik

Z k k

Y jk aij X ik
i 1

Y
l k

j 1,l

Y j ,k j , k

Zk, Xik, Yjk = 0,1 The objective function is to Minimize

Z
k 1

We consider 6 machines, 6 operations and 6 positions. We have 78 variables and 84 constraints. The optimal solution has Z = 2 with X41 = 1 and X64 = 1. We also have Y11 = Y21 = Y31 = Y41 = Y54 = 1 and Y64 = 1. There are several optima with 2 machines and we have given one of the solutions. Operations 1 to 4 are carried out on machine 4 and operations 5 and 6 are carried out on machine 1. The total times on the two machines are 12 and 6 respectively. The cycle time is 12.

Branching algorithm Cycle times and unidirectional flow


At level 1 we have 4 nodes. Node 1 does {1}, node 2 does {1,2}, node 3 does {1, 2, 3} and node 4 does {1, 2, 3, 4}. LB = 2, 7, 9 and 12 respectively. Since node 2 does operations 1 and 2, the minimum value for the sum of these times is on M4 with 7. Create nodes 5, 6, 7, 8 from node 1. These do operations {2}, {2,3}, {2,3,4} and {2,3,4,5}. The LB values for nodes 2, 3, 4 and 5 are 2, 3, 8, 11 respectively. Since K = 3, we can evaluate nodes 5, 6, 7, 8 for feasible solutions. If we consider node 5, the remaining operations {3, 4, 5, 6} have to be done on the third machine chosen. The solution is {1}, {2}, {3, 4, 5, 6}. The corresponding solution is infeasible because we do not have a machine that can do operations {3, 4, 5, 6}. We fathom node 5 by infeasibility.

Node 6 gives us {4, 5, 6} on the third machine. The minimum value is 8 on M6. We have a feasible solution {1}, {2,3}, {4,5,6} and machines M1, m3, M6 with Cycle Time (CT) = 8. This is an upper bound (UB) to Z and is updated. Node 6 is fathomed by feasibility. We fathom nodes 3, 4, 7 and 8 because LB UB . From node 2 we create three nodes 9, 10 and 11 where we do operations {3}, {3,4} and {3,4,5}. The LB values are 7, 6 and infinity. We get LB = infinity because we do not have two machines that can do {1,2} and {3, 4, 5}. We fathom node 11 because LB = infinity. We can evaluate node9. The operations set are {1,2}, [3}, {4,5,6} with CT = 8. Since the UB has not improved we do not update.

We evaluate node 10. The operations set are {1,2}, [3, 4}, {5,6}. The corresponding machines are M4, M5, M3 or M4, M5, M6 with CT = 7. Since the UB has improved we update UB = 7.
We fathom node 10 by feasibility. There is no more active node. The algorithm terminates giving the optimum solution with a cycle time of 7 units considering 3 machines.

Multiple products
1 1 2 3 4 5 3 3 4 2 2 1 2 3 1 3 3 5 3 2 3 4 4 1 2 5 6 4 3 5

5
1 1 2 5 3 4 4 5 4

2
3 4 5 6

3
3 5 4 2

4
3 1 4 5 2

3
6 3 2

The integer programming formulation has 50 variables and 69 constraints. The optimum solution is given by X1 = X2 = X6 = 1 with Z = 105. We use machines 1, 2 and 6. We also have Y11 = Y22 = Y23 = Y64 = Y25 = Y16 = 1 indicating that operations 1 and 6 go to machine 1, operations 2, 3 and 5 go to machine 2 and operation 4 goes to machine 6. The cycle times in machines 1, 2 and 6 are 6, 5, and 2. The cycle time is 6. For the second product, Z21 = Z12 = Z63 = Z64 = Z25 = 1 indicating that operations 1 and 5 go to machine 2, operation 2 goes to machine 1 and operations 3 and 4 go to machine 6. The cycle times in machines 1, 2 and 6 are 5, 6, and 6. The cycle time is 6.

S-ar putea să vă placă și