[1]
where is unknown positive definite symmetrical inertia matrix of spacecraft.
Google Scholar
[2]
is spacecraft angular velocity vector in body coordinate relative to inertia coordinate , is spacecraft control torque vector. is skew symmetric matrix.
Google Scholar
[3]
In order to avoid the singular question of large angle maneuver described with Eulerian angle, the quaternion is used to represent spacecraft attitude. By rigid-body kinematics, reference frame only needs to rotate one time from initial position to target frame. A quaternion consisted of vector and scalar are employed to represent the rotation from one coordinate to other through some axis or vector.
Google Scholar
[4]
where is identity matrix. ( is dimensional Euclidean space) is actual quaternion which body frame is relative to inertia frame, satisfies normal constraint equation , where . is defined to be skew symmetric matrix.
Google Scholar
[5]
The Eulerian angles that spacecraft rotates principal axis of inertia are recorded as roll angle , pitch angle , yaw angle . As the time of agile spacecraft maneuver is quite short, the orbit angular velocity can be neglected, then the switching relationship between quaternion and Eulerian angles can be written as following.
Google Scholar
[6]
In the mission of spacecraft attitude maneuver, there is a fixed angle between the expected frame and inertia frame. The angle can be indicated by command quaternion , the command quaternion also satisfies the constraint . The error quaternion of attitude maneuver indicates the difference between actual quaternion and command quaternion . The relationship among them is.
DOI: 10.1201/9780429446580-12
Google Scholar
[7]
where represents quaternion multiplication. Additionally, error quaternion can be expressed as.
Google Scholar
[8]
The derivation of Eq. 8 is.
Google Scholar
[9]
The Design of Controller According to the kinematics described by Eq. (4), the following Lyapunov function is selected to stabilize the subsystem.
Google Scholar
[10]
On the basis of the constraint condition.
Google Scholar
[11]
To differentiate Eq. (11), there is.
Google Scholar
[12]
For the sake of Lyapunov stability, namely keeping , the following design steps are taken. With regard to the kinematics Eq. (9), the is taken as virtual control input, the stabilization function is designed to stabilize the kinematics system of Eq. (9). New variable is defined as following.
Google Scholar
[14]
The nonlinear tracking function is proposed as[] K. S. Kim, and Y. Kim. Robust Backstepping Control for Slew Maneuver Using Nonlinear Tracking Function. IEEE Transactions on Control Systems Technology, Vol. 11, no. 6, p.822–829, 2003. ].
DOI: 10.1109/tcst.2003.815608
Google Scholar
[15]
where , are positive constant. Taking the Eq. (13) Eq. (14) into the Eq. (12), there is.
Google Scholar
[16]
Taking Eq. (15)into Eq. (16).
Google Scholar
[17]
When , , combining with character of the arctan function there is.
Google Scholar
[18]
Consequently is asymptotic stability. To differentiate Eq. (14).
Google Scholar
[19]
where can be derived from the derivation of equation Eq. (15).
Google Scholar
[20]
The new variable is defined as.
Google Scholar
[21]
Both side of Eq. (19) are left multiplied by , then bringing Eq. (1) Eq. (21)in, there is.
Google Scholar
[23]
So Eq. (22)can be written as.
Google Scholar
[24]
where is unknown constant, representing the spacecraft actual moment of inertia.
Google Scholar
[25]
The expression of is as following.
Google Scholar
[26]
The estimation error of spacecraft moment of inertia is defined as.
Google Scholar
[27]
where is estimation value of moment of inertia, because is unknown constant, the derivation of Eq. (27)is.
Google Scholar
[28]
The system Lyapunov function can be derived by augmenting Eq. (10).
Google Scholar
[29]
where is estimation matrix of the moment of inertia, to differentiate Eq. (29).
Google Scholar
[30]
There is by Eq. (27), bringing , Eq. (24), Eq. (28)into Eq. (30)yields.
Google Scholar
[31]
Obviously.
Google Scholar
[32]
Bringing Eq. (17) Eq. (32)into Eq. (31), there is.
Google Scholar
[33]
In order to ensure , the adaptive controller is designed as following.
Google Scholar
[35]
where is positive constant. Bring Eq. (34), Eq. (35)into Eq. (33)yields.
Google Scholar
[36]
Obviously, the system is global asymptotic stability. Simulation Analysis With Matlab/Simulink, simulation is carried out according to the following parameters [] N. Yan. Research on the Simulation and Experiment of Attitude Control Method for Agile Small Satellite. Beijing, Beihang University, 2011. ] in order to validate the effectiveness and superiority of the designed nonlinear backstepping adaptive controller. Spacecraft initial attitude angle . Spacecraft initial angle velocity . Spacecraft moment of inertia The estimation matrix of spacecraft moment of inertia Command quaternion . Control constants , ,. The nonlinear backstepping adaptive controller represented by Eq. (34) Eq. (35)is used to simulate, the simulation results are shown in Fig. 1-5. Fig. 1 gives the command torque time history. Fig. 2 shows the attitude angle time response. Fig. 3 is the attitude angle velocity time response. Fig. 4 depicts the quaternion time response. Fig. 5 describes the tracking error time response, and Fig. 6 draws the time response of spacecraft estimation moment of inertia. Fig. 1. The time history of command torque. Fig. 2. Time response of attitude angle. Fig. 3. The angle velocity time response. Fig. 4. Quaternion time response. Fig. 5. Tracking error time response. Fig. 6. The time response of estiation of inertia. From the aforesaid simulation results, the designed nonlinear backstepping adaptive controller fast reduced the peak of control torque, achieved the estimation of unknown moment of inertia, and ensured the closed loop stability of system. Conclusion The spacecraft large attitude maneuver based on nonlinear backstepping adaptive is studied in this paper. The non-singular quaternion is used to represent the mathematical model of spacecraft attitude maneuver. Aimed at the uncertainty of spacecraft moment of inertia, the nonlinear backstepping tracking function is adopted to design the adaptive controller both of control torque and the estimation of unknown moment of inertia, the system stability is analyzed by Lyapunovo function. The simulation study is carried out through Matlab/Simulink. The simulation results show that the designed nonlinear backstepping adaptive controller can fast reduce the peak of control torque, achieve the estimation of unknown moment of inertia, and ensure the closed loop stability of system. The controller is validated to be effective and feasible. Acknowledgement This work partially supported by National Natural Science Foundation and Youth Science and Technology Foundation. Reference.
Google Scholar