This is my first question here. I am working on a binary LP wherein I want to turn on a variable when a condition is met else make it zero. I want to share my formulation with you and need your help in understanding whether my constraint gets the job done.
My main motive is to express the following indicator constraint using the big-M formulation. $$f^{t}_{ls} = 1 \qquad if \sum_{c\in C}f^{t,c}_{ls} \geq d_t\qquad \forall t\in T, ls\in L_s\\ 0\qquad\qquad else\qquad\qquad\qquad\qquad $$
My objective is to minimise the costs incurred $$\textbf{min}\qquad \sum_{c\in C}\sum_{t\in T}\sum_{ls\in L_s}cost(c,ls)f^{t,c}_{ls}$$
To express the indicator constraint in the big-M version, I re-wrote this as follows:
$$\sum_{c\in C}f^{t,c}_{ls} + M(1-f^{t}_{ls}) \geq d_t \qquad \forall t\in T, ls\in L_s$$ where both variables $f^{t,c}_{ls} and f^{t}_{ls}$ are binary and $d_t\in Z^+$.
Could you please let me know if this works? I hope I didn't get anything wrong
Thanks a lot