Limit Cycle Analysis and Suppression for Disturbance Observer-Based Missile Autopilot With Flight Test Results

Zonghua Sun, Liaoni Wu, Yancheng You

Abstract: Limit cycle is a typical phenomenon in the nonlinear system, usually deteriorating stability even damaging the systems. This article investigates the formation mechanisms and suppression method of the limit cycle oscillations induced by the actuators backlash nonlinearity under the disturbance observer-based autopilot system. First, the forming reason for the high gain nature of disturbance observers is interpreted from the nonlinear dynamic inversion perspective. Subsequently, a graphical analysis approach that integrates the describing function technique with transfer function representations is presented to predict the limit cycle frequency, amplitude, and stability. The effects of the backlash parameters and controller gains on the limit cycles are given. Finally, an adaptive tuning mechanism of the observer gains only in terms of tracking error is proposed to attenuate the limit cycle oscillation caused by the interaction between unmodeled backlash nonlinearity and high observer gain nature. Numerical simulations and flight test results indicate that the proposal protects the system from a periodical limit cycle with high tracking accuracy, strong robustness, and effective disturbance attenuation.

External IDs:dblp:journals/taes/SunWY25