Research On Strong Robust Attitude Control With Input Constraints For Near Space Vehicles During Reentry

Considering the tense international situation and the further requirement of national defensesecurity, the near space vehicle (NSV) has become a research focus of the military powers, and it hasimmeasurable strategic value in the military and civilian fields. The NSV flies in the near space whichis a link between the aviation and aerospace. The reentry attitude control system is an important partof the flight control system of the NSV and it is also the foundation of the safe return after the mission.But the problem of constrained inputs, complex flight environment and large external momentdisturbance makes the design of the reentry control system more complicated. Around these problems,the thesis conducts the corresponding researches:First, the NSV reentry model is established. The fast and slow loop attitude controllers aredesigned based on global fast terminal sliding mode control method, in view of strong nonlinearity,strong coupling and fast time-varying characteristics. Then the attitude states of the vehicle canconverge to equilibrium states in a limited time. Additionally, the symbol function is replaced with thesaturation function, which reduces the chattering effectively.Second, the NSV reenters from the high airspace where the air is thin. So the control surfacesmust deflect large enough to control the NSV. Aiming at the problem of input constraints duringreentry, external anti-windup compensation system is proposed to eliminate the influence on thesystem and the control compensation for constrained inputs is realized. Through it, the attitude anglescan track the command signals fast and smoothly.Finally, aiming at the problem of model parameter uncertainty and external moment disturbanceof the NSV, the adaptive sliding mode disturbance observer based on super-twisting algorithm isproposed to estimate the uncertainty/disturbance and the compensation control law is designed. Themethod, without the derivative of disturbance and with less learning parameters, is suitable for thereal-time control. And the effectiveness of this method is demonstrated through the simulation test.