Nonlinear Model Predictive Control for Near-Space Interceptor Based on Finite Time Disturbance Observer

This paper proposes a novel nonlinear model predictive control design method based on the finite time disturbance observer (FTDO) and dynamic control allocation (DCA) algorithm for the near-space interceptor (NSI). First, the FTDO is introduced to estimate the mismatched system disturbance. The nonlinear model predictive control scheme based on the estimation value is proposed to obtain the virtual control command, and the mismatched disturbance can be effectively removed from the output channels of the NSI by designing a feed-forward disturbance compensation term with an appropriate compensation gain matrix. Moreover, a new DCA method is given to distribute the above virtual control command among the corresponding actuators (reaction jets and aerodynamic fins). Finally, numerical simulations illustrate that the proposed control scheme can track the command signal with high precision and have strong robustness against the mismatched disturbance.