Model Predictive Control

Why MPC?

  • Very good method for constraints handling
  • Systems with complex dynamic can benefit from considering this already in the computation of the control inputs
  • Intuitive method for control of MIMO systems

The MPC work flow

In class, we are taught mainly about how to choose suitable terminal sets and weights, how to prove stability, and other more or less theoretical aspects. However, this is only a fraction of the work that acctually goes into the MPC design process:

Figure: Step response for the differential equation.