I'm trying to develop a heuristic to solve a scheduling problem. The difficult for me is to treat the time as a variable.
There are 2 variables related to time, $b_{ij}$ and $f_{ij}$, being the time that the machine j begin and end the job $i$ respectively. It's used like this because the model has a demand constraint (see below), that is attended with the time that the machines do job $i$, and each machine has a specific capacity ($q_j$). There is also in the model a binary variable $x_{ij}$ that says if the machine $j$ do the job $i$.
$$\sum_j \bigg((f_{ij} - b_{ij}) * q_j * x_{ij}\bigg) \geq Demand_i, \forall i$$
In this problem, all the jobs needs to be processed only one time by whatever machine.
If anyone knows which kind of scheduling problem it is... It'll be a great help.