Volatile electricity prices make demand response (DR) attractive for the process industry. However, the resulting scheduling optimization problems are notoriously challenging to solve, especially in the case of nonlinear process dynamics. Thus, problem reformulations that allow online scheduling are needed. In our previous publication (arXiv:2110.08137), we proposed high-order dynamic ramping constraints with non-constant ramp limits derived rigorously for input-state linearizable single-input single-output (SISO) processes. In the present publication, we generalize dynamic ramping constraints to flat multi-input multi-output (MIMO) processes. Approximating the nonlinear ramping limits by conservative piecewise-affine functions allows to formulate a mixed-integer linear programming (MILP) optimization problem. Moreover, in the MIMO case, dynamic ramping significantly reduces the model dimension. As a case study, dynamic ramping constraints are derived for a heated reactor-separator process with recycle that is subsequently scheduled simultaneously with a multi-energy system. The dynamic ramping formulation bridges the gap between rigorous process models and simplified process representations that allow online scheduling.