Sustainable multi-commodity capacitated facility location problem with complementarity demand functions

We investigate a multi-commodity capacitated facility location problem involving sustainability concerns (e.g., restrained carbon emission). In addition, this problem incorporates a complementarity demand function. We show that this incorporation can lead to superior decisions both analytically and computationally. The resultant formulation is a 0-1 mixed-integer non-concave quadratic program with equilibrium constraints. We adopt the piecewise-linear envelope method to transform the formulation into a 0-1 mixed-integer concave program. We then propose an efficient branch-and-refine algorithm with global convergence. Numerical examples demonstrate the effect of carbon emission limit and carbon trading on company decisions.