A velocity algorithm for the implementation of gain-scheduled controllers (Q1899611)

Gain-scheduled controllers are effective means of controlling performances of dynamical systems which operate over a wide range of operating conditions. This paper presents a velocity (D-method) algorithm for implementation of gain-scheduled controllers for nonlinear plants. The need to differentiate some measured outputs, which may pose noise amplification problems, is indeed a weakness of this method. Also the global stability is difficult to guarantee due to the time-varying nature of the gain-scheduling operations. Before the stability issue is properly addressed, such a design method can at best be considered as an ad hoc procedure. Although it is difficult to tell how large the class of systems is to which this method is applicable, nevertheless this design method should be useful to some practical systems. The paper would be more interesting if some numerical examples and simulation results were included to illustrate its merits. More results concerning the stability of time-varying systems will be needed to make gain-scheduled controller design a formal design method.