I have to optimize this (see image
) but I have to do it with respect to lambda (use of machines) and b (batch sizes), I'm using intellij idea using the cplex library, but I get an error: CPLEX Error 5002: 'q1' is not convex , I don't know how to solve ..
for this reason I was thinking otherwise of calculating first with respect to lambda and instead of b I insert the maximum capacity value (value already given), so once I found lambda I replace the values and then I optimize with respect to b. Does this make sense? I'm doing it wrong? what would be the right method to solve?
obviously here the constraint is missing that the sum of the lambda on the m -> machines must be = 1 for each product i.
I have to do is the third point
Asked
Active
Viewed 47 times
0
Fernanda Rodríguez
- 21
- 3
-
I would try to linearize by discretizing λ/b. – Erwin Kalvelagen Dec 03 '19 at 20:19
-
how can i do this? – Fernanda Rodríguez Dec 08 '19 at 20:47
-
1(1) simplify `min 1/f` to `max f`. (2) write y=λ/b as yb=λ . Expand b as a sum of binaries B(i). Linearize y B(i) using standard techniques. – Erwin Kalvelagen Dec 08 '19 at 21:04