TABLE OF CONTENTS The approach taken in documenting this new controls design methodology is as follows. The method is first briefly summarized. A simple example is then covered to give the reader a flavor for the design algorithm.58 This example shows how the method is easily used to generate a frequency and amplitude dependent H ? controller for a problem that cannot be solved using other existing techniques, such as dynamic inversion.67 This method is then used on a more realistic design problem, the roll axis autopilot for a UAV (unmanned aerial vehicle).68 The descriptions are subdivided into the three major steps of the design process. They are13
Generation of a describing function for the nonlinear system.
Minimizing the H ? performance index in the frequency domain for the describing function of the plant.
Generation of the stability domain for the closed-loop system using computational geometry. If desired, the regions of stability can also be generated for the open-loop system or for the isolated controller.
The three-step process covered in this design approach can be summarized as follows: 1) modeling, 2) synthesis, and 3) analysis.13 Each of these steps corresponds to a specific technique, which is blended into a cohesive methodology taking the controls engineer from the initial problem description to the final verification of the controls solution.
The plants covered by this methodology are nonlinear, which is important as all systems in nature are intrinsically nonlinear. The degree of the nonlinearity determines...
