I am trying to bring logic and programming to a currently manually driven process.
We take the weights of 16 different crushing hammers, organize them into sets of 4 based on how close each set's weight is to the others. We are looking to have less than 1 pound difference between all 4 sets. The weights are known but I cannot logically program a method to do this without pen and paper.
Example below.
| Set A | Set B | Set C | Set D | |
|---|---|---|---|---|
| 39.1 | 40.1 | 42.0 | 41.5 | |
| 40.05 | 41.0 | 40.05 | 38.90 | |
| 41.2 | 42.1 | 41.3 | 43.1 | |
| 38.5 | 43.60 | 42.1 | 41.5 | |
| Totals | 158.85 | 166.80 | 165.45 | 165 |
As you can see in the first example, Sets C and D are close enough. Sets A and B are too far apart and need readjusting, but due to the severe difference, I would most likely have to rearrange all 4 sets to get with 1 pound difference for all 4 sets. Now mind you, this is all done on paper, and I am looking for a way to plug in all numbers and have it spit out the best configuration given the hammer weights, no more paper.
| Set A | Set B | Set C | Set D | |
|---|---|---|---|---|
| 42.1 | 39.1 | 42.0 | 41.5 | |
| 40.05 | 41.0 | 40.05 | 38.90 | |
| 43.60 | 43.1 | 41.3 | 42.1 | |
| 38.5 | 41.20 | 40.1 | 41.5 | |
| Totals | 164.25 | 164.40 | 163.45 | 164 |
I could do this all day, splitting hairs to get as close as possible. The closer we get the weights, the less vibration we experience and our equipments last a lot longer. Anyone have any thoughts to accomplishing this?
I'm not a good math guy but I think in order to arrange ALL 4 SETS within 1lb tolerance, all 16 hammers should have some what consistent weight that is a bit apart from targeted weight.
The most simple way I can think of is to sort all of them in order and assign index 0~3 to each group until all 16 are assigned.