Controller design for nonlinear multivariable systems

A new method of controller design for use with nonlinear multivariable systems with discontinuous nonlinear terms is developed. The procedure is based on generating the sinusoidal-input describing function models of the plant followed by application of a nonlinear optimization technique. The procedure is applied to bank-angle control of a highly nonlinear unmanned aerial vehicle (UAV), and the results are compared with three other controllers that were previously designed for the same system. It is concluded that the present controller out-performs a linear controller, and it competes with a nonlinear controller.