Comparison among gap-based weighting for multi-model control of a TITO system

The gap metric has been employed to formulate weighting functions for multi-model control of nonlinear systems. Compared to traditional weighting method, there is only one tuning parameter, so complicated tuning task is avoided. Besides, the gap metric based weighting methods extend the merits of traditional ones: they can also be computed offline and stored in lookup tables. Comparison among three kinds of gap metric based weighting method has been made in this work. A Two-input-two -output (TITO) system is studied. Simulations demonstrate that each of the gap metric based weighting method has its typical advantage. And they are all superior to traditional weighting methods.