This project considered the relative performance between two standard classes of robust adaptive controllers, namely designs based on the dead-zone and projection modification.
Simple criteria were established to indicate when each design outperformed the other w.r.t. to a non-singular performance cost functional penalising both the output and the control transient.