Synthesis of missile gain-scheduled autopilots using an $H_\infty$-LPV technique with piecewise continuously differentiable parameter dependent Lyapunov functions

  • P.C. Pellanda
  • A.M. Simoes
  • P. Apkarian
  • D. Alazard

Abstract

The missile pitch-axis autopilot design is reconsidered using a classical result in Linear Parameter-Varying (LPV) control. We consider a nonlinear missile model with a highly nonlinear dependence on the scheduling variables on which the application of traditional finite-dimensional LPV methods as the LFT- or polytopic-based techniques is difficult or conservative. The main objective is to obtain gain-scheduled autopilots that guarantee closed-loop $L_2$-gain performances with little conservatism by providing a general LPV/gridding technique state) Lyapunov functions. An iterative and computationally feasible procedure is proposed to construct functional dependencies for the Lyapunov variables that are rich enough to ensure a suitably specified performance level over a wide range of flight conditions. Gain-scheduled autopilots are synthesized using the proposed approach and results are compared with different LPV synthesis strategies.
Published
2004-05-01