Abstract

Shimmy oscillation is an anxious concern in aircraft landing gear design and maintenance. Through related literature review, it is found that active shimmy control (suppressing) is still an open problem. In this thesis, an in-depth analysis has been carried out on aircraft landing gear shimmy dynamics and active control strategy has been developed to suppress shimmy oscillation. Based on a nominal aircraft landing gear model, its shimmy Limit Cycle Oscillation (LCO) variation with respect to varying parameters has been studied by numerical integration. The shimmy stability variation with varying caster length and taxiing velocity has also been analyzed after linearizing the system. Due to inherent system dynamics uncertainties (such as varying taxiing velocity) and external disturbances (such as rough runway), Robust Model Predictive Control (RMPC) technology is resorted to suppress shimmy during aircraft landing. A new active control strategy has been proposed suitable for online shimmy control application by combining a RMPC control law with a LPV polytope design. The proposed RMPC has been compared with two present RMPCs both in a benchmark example and the landing gear shimmy control. It has been verified by simulation results that the proposed RMPC stabilizes the unstable parameter-varying landing gear system with guaranteed closed-loop stability, high computational efficiency and strong disturbance rejection ability. Related design parameters, mathematical proof and implementation considerations are addressed in this thesis.