“…• Linear programming [2,5,17] and network flow [25] • Convex nonlinear optimization [6,26,27,31], including Lagrangian relaxation [4,6,19,30,31] • Slew budgeting [14] Due to the complexity of modern designs, the two most scalable approaches are those based on Lagrangian relaxation and sensitivity. Instead of satisfying every imposed constraint, Lagrangian relaxation changes the original constrained problem into an unconstrained problem such that the solutions to the latter can be mapped back to the former.…”