You are on page 1of 22

Author’s Accepted Manuscript

Thrust Vector Control of Upper Stage with a


Gimbaled Thruster during Orbit Transfer

Zhaohui Wang, Yinghong Jia, Lei Jin, Jiajia Duan

www.elsevier.com/locate/actaastro

PII: S0094-5765(16)30526-4
DOI: http://dx.doi.org/10.1016/j.actaastro.2016.06.002
Reference: AA5848
To appear in: Acta Astronautica
Received date: 7 February 2015
Revised date: 22 May 2016
Accepted date: 1 June 2016
Cite this article as: Zhaohui Wang, Yinghong Jia, Lei Jin and Jiajia Duan, Thrust
Vector Control of Upper Stage with a Gimbaled Thruster during Orbit Transfer,
Acta Astronautica, http://dx.doi.org/10.1016/j.actaastro.2016.06.002
This is a PDF file of an unedited manuscript that has been accepted for
publication. As a service to our customers we are providing this early version of
the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting galley proof before it is published in its final citable form.
Please note that during the production process errors may be discovered which
could affect the content, and all legal disclaimers that apply to the journal pertain.
Thrust Vector Control of Upper Stage with a Gimbaled Thruster
during Orbit Transfer

Zhaohui Wang1, Yinghong Jia1*, Lei Jin1, Jiajia Duan2


(1 Beihang University, School of Astronautics, Beijing, 100191, P.R. China;
2 Shanghai Institute of Aerospace Control Technology,Shanghai, 201109,China)

Abstract: In launching Multi-Satellite with One-Vehicle, the main thruster provided by the

upper stage is mounted on a two-axis gimbal. During orbit transfer, the thrust vector of this

gimbaled thruster (GT) should theoretically pass through the mass center of the upper stage and

align with the command direction to provide orbit transfer impetus. However, it is hard to be

implemented from the viewpoint of the engineering mission. The deviations of the thrust vector

from the command direction would result in large velocity errors. Moreover, the deviations of

the thrust vector from the upper stage mass center would produce large disturbance torques. This

paper discusses the thrust vector control (TVC) of the upper stage during its orbit transfer. Firstly,

the accurate nonlinear coupled kinematic and dynamic equations of the upper stage body, the

two-axis gimbal and the GT are derived by taking the upper stage as a multi-body system. Then,

a thrust vector control system consisting of the special attitude control of the upper stage and the

gimbal rotation of the gimbaled thruster is proposed. The special attitude control defined by the

desired attitude that draws the thrust vector to align with the command direction when the gimbal

control makes the thrust vector passes through the upper stage mass center. Finally, the validity

of the proposed method is verified through numerical simulations.

Keyword: Upper stage, Thrust vector control, Orbit transfer

*Corresponding author.
Email address: jia_yingh@163.com

1
I. Introduction

THRUST vector control has been considered as a crucial area of research in the launch
systems, because of the extremely advantageous ability to control the thrust vector of any
[1-5]
propulsion system. For the launch systems , the direction of thrust vector of the propulsion

system should align with the command direction. The TVC of the launch systems can be

obtained by gimbaling the engine via servo actuation system. The deviations of the thrust vector

in the launch systems lead to large velocity and positioning errors, along with extra requirements

of orbit control. For the upper stage, it should transfer to the other orbit when it completes the

deploy mission in one orbit. During orbit transfer, the thrust vector should theoretically align

with the command direction and pass through the mass center of upper stage. The thrust vector

misalignments of the upper stage also produce disturbance torques and make the upper stage

rotate. Furthermore, it results in long orbit transfer time and extra requirements of attitude/orbit

control.

TVC has become a significant technology for the launch systems to control their trajectories

and accomplish high maneuverability. The requirements for the TVC of space shuttles were

presented in Penchuk’s work [1]. The TVC for space shuttles obtained by Omni-axis vectoring

of nozzle was studied in reference [2]. The controller can maintain the vehicle trajectory and the

direction of the thrust vector by independently controlling the gimbal angles of the engine. The

model, estimation, implementation and control techniques of the TVC-electromechanical

actuators for the rocket engines were investigated in references [3-5]. Taking fuel slosh dynamics

into consideration, Jaime [6] researched on the TVC for the rocket engine. However, three more

independent torques were needed in the control system. The TVC for a liquid-bipropellant upper

stage spacecraft was presented by Wie [7]. The controller can keep the pointing of the thrust

2
vector and the stability of the spacecraft. In researches [8, 9], the TVC for a solar-sail spacecraft

was proposed. The sail angle with respect to the sunlight was utilized to produce a net torque for

the attitude control and thrust vectoring. For the satellite systems [10, 11], the TVC was used to

correct the misalignments produced by the thruster firing and the shift of the mass center for a

satellite due to fuel consumption, solar panels deployment, onboard experiments, etc. For the

upper stage, the selection of a TVC system actuation technology and subsystem technology was

discussed in Garrison’s research [12]. Most achievements above were concentrated on Linear

Actuators with small engine deflection defined in Suchitra’s study [13], where Linear Actuators

are most powered by pneumatic or hydraulic actuators, and Rotary Actuators are motor powered

by electromechanical actuators. The Linear actuators were used for small engine deflection ( 4 ),

whereas for higher gimballing of the engine; preference goes to Rotary Actuators. In this paper,

Rotary Actuators are utilized to solve the higher gimbaling case.

Disturbance torques produced by the thrust vector misalignments, eight reaction control

subsystem (RCS) thrusters, together with the main thruster, are utilized in the TVC in order to

eliminate the effects of the thrust vector misalignments. The thruster configuration presented in

this paper was defined as the optimal one in Hwang’s investigation [14]. The steering law for the

RCS thrusters is obtained by pulse-width and pulse-frequency modulator (PWPFM) which is

first utilized in the attitude controller by Bong W. [15] and has a first order filter compensates the

Schmitt trigger output in the feedback path. The advantage of the PWPFM is the static

parameters that are independent with respect to the parameters of the spacecraft. The accurate

nonlinear model of the upper stage is derived by taking the upper stage as a multi-body system

which consists of the upper stage body, the gimbaled thruster actuator and the gimbaled thruster.

The superiority of this context is the thrust vector control algorithm which completes by the

3
special attitude control of the upper stage and the gimbal rotation of the gimbaled thruster. To

achieve this goal, the algorithm that determines the desired attitude for the special attitude

control and the gimbal control method that based on the special attitude control is proposed. Due

to the property of the rotary actuator, the rotation motion of the gimbal and the GT is strongly

coupled with the attitude motion of the upper stage body. The control scheme takes the nonlinear

coupling effect into consideration because the controller based on the linearization model is

inappropriate in this situation. The remainder of this paper is outlined as follows. First, the

comprehensive mathematical formulations are proposed considering the coupling of the orbit and

attitude motion, the upper stage’s motion and GT’s rotation. Next, the desired attitude of the

upper stage is obtained using the newly derived algorithms to guarantee the direction of the GT’s

thrust vector aligning with the command direction. In the same way, the desired gimbal angles of

the GT are got to make sure the GT’s thrust vector passing through the mass center of the upper

stage. Then, a TVC system including the upper stage’s attitude control and the GT’s gimbal

control is designed, ensuring the thrust vector of the GT not only passes through the mass center

of the upper stage, but also aligns with the command direction. Finally, the numerical

simulations are performed, followed by analysis and conclusions.

II. Mathematical Models

The system discussed in this paper consists of an upper stage body, a two-axis gimbal and a

GT. The kinematic and dynamic equations of the system are derived in this section. The

configuration of the system is shown in Fig.1.

4
RCS Thruster
RCS Thruster
Upper Stage
yb
Gimbaled Thruster Upper Stage
xb Fm
ob Gimbaled Thruster
RCS Thruster
RCS Thruster RCS Thruster

R
Z
RCS Thruster RCS Thruster

Y
X

Fig. 1 Configuration of upper stage

A. Coordinate frames

To clearly represent the dynamics of the system, the coordinate frames are defined as follows.

An Earth centered inertial frame is denoted by f I , with its origin located at the center of the

Earth, the z I axis pointing towards the celestial North Pole, the xI axis pointing towards the

vernal equinox and the y I axis completes the right-handed reference frame. The orbit reference

frame ( f o ), has its origin located at the mass center of the upper stage, with the zo axis pointing

towards the Earth center, the xo axis pointing towards the direction of tangential velocity and the

yo axis completes the right-handed reference frame. The principal body-fixed coordinate frame

of the upper stage is denoted by f b with its origin located at the mass center of the upper stage.

Attitude orientation of the body fixed frame f b with respect to frame f o is represented by ( ZXY )

set of Euler angles:  (pitch) about z axis,  (roll) about x axis and  (yaw) about y axis. The

body-fixed reference frame of the gimbal is denoted by f k , and the gimbal angle of f k with

respect to f b is represented by Euler angle:  (roll) about x axis. The body-fixed reference

5
frame of the GT is denoted by f m , and the gimbal angle of frame f m with respect to frame f k is

represented by Euler angle:  (yaw) about y axis. The coordinate frames are shown in Fig. 2.

zo

zb 
Upper Stage yk  ym  z
m zk
yb 
  yb
xb 
xk Fm 
ob  yo
 fb   f k / f m  ok  om 
xm
Gimbaled Thruster 
R
xo 
Z
 fo  xb

 fI  O Y
 fb 
X

Fig. 2 Definition of coordinate frames

B. Thrusts and Torques

The thrust vector of the GT is denoted by Fm , which is demonstrated in Fig. 2 and defined as

Fm  Fm I m  0J m  0 K m  Fm cos  sin  Io  Fm sin  J o  Fm cos  cos  K o (1)

where Ii , J i , K i are unit vectors along the axes of frame f i ( i  o, b, k , m ),  is the elevation

angle,  the azimuth angle, and Fm the constant thrust level of the GT.

Jo
Fm

Oo 
Io

Ko

Fig. 3 Gimbaled Thruster azimuth and elevation

6
To calculate the thrusts and torques provided by the GT, two consecutive gimbal rotations are

utilized. First, the two-axis gimbal rotates with respect to the upper stage body about the xk axis,

 is the gimbal angle. Second, the GT rotates with respect to the two-axis gimbal about the

intermediate axis y ' ,  is the gimbal angle. Then, Fm expressed in frame f b is

Fm  Fm cos  Ib  Fm sin  sin  Jb  Fm sin  cos  Kb (2)

rm  rmx Ib  rmy Jb  rmz Kb is the vector presenting the placement from the upper stage’s mass

center to the action point of the GT. Then, the torques provided by GT is

Tm  rm  Fm (3)

In this paper, we assume that all thrusters can provide constant thrusts. The torque provided

by the RCS thrusters is expressed as

Tx  4Fg rg , Tx  4Fg rg (4)

where Tx , and Tx represents the positive and negative torque along xb axis respectively, Fg the

constant thrust level of the RCS thrusters, and rg the moment arm.

C. Kinematic Equations

Based on the definition of the attitude orientation for the upper stage, the kinematic equation

gives

   rx cos   rz sin  


   
σ     ry  tan   rx sin   rz cos     A  ω  Abo ωo  (5)
   ( sin    cos  ) / cos  
   rx rz 

where σ      is the attitude of the upper stage, A the transformation matrix from
T

T
ωr to σ , ωr  rx ry rz  the angular velocity of the upper stage body with respect to

7
T
frame f o , ω  x  y z  the angular velocity of the upper stage body with respect to

frame f I , A ji the transformation matrix from frame f i to frame f j ( i, j  o, b, k , m ) and ωo the

orbit angular velocity.

The rotating equation for the gimbal with respect to the upper stage body is

ωk  Γ k (6)

where ωk is the angular velocity of the gimbal with respect to the upper stage body expressed in

frame f k . Γ k  [1 0 0]T is determined by the configuration of the system.

The rotating equation of the GT with respect to the gimbal gives

ωm  Γ m  (7)

where ωm is the angular velocity of the GT with respect to the gimbal expressed in frame f m .

Γ m  [0 1 0]T is determined by the configuration of the system.

D. Dynamic Equations

The dynamic equations including both translational and rotational motion of the system are

derived based on Kane equation [16]. The general formula of Kane equation is defined as

Fi G  Fi A  0 (8)

where Fi G , Fi A are the generalized inertia force and generalized active force of the system with

respect to the ith order generalized velocity.

First, the dynamic equation for the translational motion of the upper stage is

mvb  Sbt ω  Abk Skt Γ k  Abm Sm Γ m   Fb,n  Fb  Fw (9)

where

Fb,n  mωvb  ωSb*ω  Abk Ωk Sk* Ωk  Abm Ωm Sm Ωm  Abk Skt Akb ωAbk Ωk  Abm Sm Amk Ωk Akm Ωm

8
where m  mb  mk  mm , mb , mk , mm are the masses of the upper stage, the two-axis gimbal and

the GT respectively, vb is the velocity of the upper stage defined in frame f b , Sbt the static

moment of the system expressed in frame f b , S kt the static moment of the gimbal and the GT

expressed in frame f k , S k the static moment of the gimbal expressed in frame f k , S m the static

moment of the GT expressed in frame f m , Sb the static moment of the upper stage’s augment

expressed in frame f b , S k the static moment of the gimbal’s augment expressed in frame f k ,

Ωk the angular velocity of the gimbal, Ωm the angular velocity of the GT, Fb the thrust and Fw

the gravity that act on the system.

The dynamic equation for the rotational motion of the upper stage is

 SbtT vb  Ibt ω  Abk I kb Γ k  Abm I mb Γ m   Tb,n  Tb (10)

where

Tb,n  Sbt ωvb  I ω  Abk Ib k Ωk  Abm Ib m Ωm  Abk Ikb Akb ωAbk Ωk  Abm Imb Amk Ωk Akm Ωm

where I bt is the inertia matrix of the system expressed in frame f b , I kb the coupling inertia

matrix of the gimbal, I mb the coupling inertia matrix of the GT, I the quasi inertia matrix of the

system, I b k the quasi inertia matrix of the gimbal, I b m the quasi inertia matrix of the GT and Tb

the torque which acts on the system.

The dynamic equation of the gimbal rotates with respect to the upper stage body is

 
Γ kT  SktT Akbvb  ( I kb )T Akb ω  I kt Γ k  Akm I mk Γ m   Tk ,n  Tk (11)

where

Tk ,n  Skt Akb ωvb  SktT Akb ωrb,k ω  Akm Ik m Ωm  Ikt Akb ωAbk Ωk  I k Ωk  Akm Imk Amk Ωk Akm Ωm

9
where I m denote the inertia matrix of the GT, I mk the coupling inertia matrix of the GT, I Ωk the

quasi inertia matrix of the gimbal, I Ωk m the quasi inertia matrix of the GT, and Tk the rotational

control torque which acts on the gimbal.

The dynamic equation of GT rotates with respect to the gimbal is

 
Γ mT  SmT Amb vb  ( I mb )T Amb ω  ( I mk )T Amk Γ k  I m Γ m   Tm,n  Tm (12)

where

Tm,n  Sm Amb ωvb  SmT Amb ωrmω  Ωm Im Ωm  ( Imk )T Amb ωAbk Ωk  Im Amk Ωk Akm Ωm

where Tm is the rotational control torque that acts on the GT.

Obviously,Eq. (9) is coupled with Eqs. (10), (11) , and (12). For TVC, Eqs. (10), (11) and

(12) are independent. Substituting Eqs. (5) and (9) into Eqs. (10), (11) and (12), the independent

dynamic equations can be rewritten as

At x  T  Tn (13)

T T
where x  σ T    , T  TbT Tk Tm  . The elements of the matrix At and the nonlinear
thrust Tn are given as

  1 T  1  1 T   1 T  
 I bt  Sbt Sbt  A  Abk I k  Sbt Abk Skt  Γ k  Abm I m  Sbt Abm Sm  Γ m 
b b

  m   m   m  
  1  1  1   1  
At   Γ kT  ( I kb )T Akb  SktT Akb Sbt  A Γ kT  I kt  SktT Skt  Γ k Γ kT  Akm I mk  SktT Akm Sm  Γ m 
  m   m   m  
 T b T 1 T  1  1   1  
 Γ m  ( I m ) Amb  Sm Amb Sbt  A Γ mT  ( I mk )T Amk  SmT Amk Skt  Γ k Γ mT  I m  SmT Sm  Γ m 
  m   m   m  

10
 1 
 I bt  SbtT Sbt 
 Tb,n   SbtT   m 
 T  1 T T   T b T   1
m T T  m 

Tn    Γ k Tk ,n    Γ k Skt Akb   Fb  Fw  Fb ,n    Γ k  ( I k ) Akb  Skt Akb Sbt   A A A
1 T 1
σ  Abo ωo 
 Γ mTm,n     
 Γ m Sm Amb 
T
 T b T 1 T 
 Γ m  ( I m ) Amb  Sm Amb Sbt  
  m 

III. Control Law

This section discusses the TVC system including the attitude control of the upper stage and

the gimbal rotation control of the GT. First, the desired gimbal angles  d ,  d  , which ensure the

thrust vector of the GT passes through the mass center of the upper stage, are given. Then, the

desired attitude of the upper stage denoted by σ d is calculated; guaranteeing the direction of the

thrust vector of the GT denoted by  ,   aligns with its command direction  d ,  d  . The

methods and analysis tools of variable structure control (VSC) are utilized in this paper due to

their robustness to nonlinear model errors.

A. Desired States

Assume that the thrust vector of the GT passes through the mass center of the upper stage so

that the desired gimbal angles  d ,  d  are defined as follows

Fm cos d / rmx  Fm sin d sin  d / rmy   Fm sin d cos  d / rmz (14)

The thrust vector of the GT also can be defined by two Euler angles  ,  which indicate the

rotation of GT with respect to frame f o . The first rotation is about the yaw ( zo ) axis, and  the

rotate angle. The second rotation is about the intermediate pitch axis ( y ' ), and  the rotate angle.

The transformation matrix Amo and thrust vector Fm in terms of  ,  are

11
cos  cos  cos  sin   sin  
Amo    sin  cos  0  (15)
 sin  cos  sin  sin  cos  

Fm  Fm cos  cos  I o  Fm cos  sin  J o  Fm sin  K o (16)

The rotation angles  ,  can be calculated by the azimuth angle and the elevation angle

 ,   as

  arctan  tan  sin  



 (17)
   arcsin  cos  cos  

The transformation matrix Amo can also be defined by  σ,  ,   as

Amo  Amb Abo (18)

where
cos  sin  sin   sin  cos  
Amb   0 cos  sin  
 (19)
 sin   cos  sin  cos  cos  

cos  cos  sin  sin  sin cos  sin  sin  sin  cos  sin  cos  
Abo    cos  sin cos  cos sin  
 (20)
sin  cos  cos  sin  sin sin  sin  cos  sin  cos cos  cos  

First, the command direction  d ,  d  of the thrust vector for the GT is determined by the

guidance system. Then, the desired rotation angles  d ,d  , the desired transformation matrix

d
Amo , the desired gimbal angles  d ,  d  , and the desired transformation matrix Amb
d
can be

calculated by Eqs. (17), (15), (14) and (19), respectively. Finally, the desired transformation

matrix Abod is obtained as

Abod   Amb  Amod


d 1
(21)

12
Thus, the desired attitude σ d can be obtained from Eqs. (20) and (21).

B. Design of Control Law

For a TVC system, the attitude control signals are the current desired gimbal angles  cd , cd 

providing the attitude control torque about yb axis and zb axis, and the on-off thrust of the RCS

thrusters providing the attitude control torque about xb axis. The GT’s gimbal rotation control

signal is the control torque provided by the electrical motor.

The methods and analysis tools of VSC are utilized in this paper due to their robustness to

nonlinear model errors. If the desired system states are defined as xd  σd cd cd  , a
T

sliding surface is defined as

S  e  k1e (22)

55
where e  x  xd is the error of system states, k1  is the weighing matrix for e .

Substituting Eq. (13) into the differentiating Eq. (22), yield

S  e  k1e   x  xd   k1  x  xd   At1 T  Tn   xd  k1  x  xd  (23)

The approaching law is chosen as

S   KS  ε sgn  S  (24)

where K  55
, ε 55
, and sgn   is the sign function defined as

1 if x  0

sgn  x   0 if x  0
1 if x  0

Substituting Eqs. (24) and (13) into Eq. (23), we obtain


T  Tn  At xd  Kk1  x  xd   ε sgn  x  xd   k1  x  xd    K  k1  x  xd   (25)

13
To determine the current desired gimbal angles of the GT  cd , cd  , substituting Eq. (25)

into Eq. (3), yield

rmz Fm cos cd  rmx Fm sin cd cos  cd  T  2 



 (26)
rmy Fm cos  cd  rmx Fm sin cd sin  cd  T  3

where T  2  , T  3 represent the second and third element of control torque (Eq. (25)),

respectively. The current desired gimbal angles  cd , cd  can be figured out by Eq. (26).

Because of the coupling, the GT would also produce torques about xb axis, which should be

counteracted by the RCS thrusters. Then, the torques that the RCS thrusters should provide is

Tbx  T 1  rmz Fm sin cd sin cd  rmy Fm sin cd cos  cd  (27)

The control torque about xb provided by the RCS thrusters is defined by Eq. (4). The steering

law of the RCS thruster is obtained by the PWPFM method [15,17,18].

In order to verify the stability of the controller, set the Lyapunov Function as V  S T S / 2 .

Substituting Eqs. (25) and (24) into the differentiation of the Lyapunov Function, yield

V  S T S   S T K T S  sgn  S T  εT S   K j S 2j    j S j  0
5 5
(28)
j 1 j 1

5 5
Since V  S T S / 2  0 , and V   K j S 2j    j S j  0 , V  0 exists only when S  0 .
j 1 j 1

From the LaSalle invariance principle, we conclude that the controller designed previously is

global asymptotic stable. The paper, in order to avoid of chattering at the sliding surface

boundary, utilizes the saturation function to instead of the sign function in the controller, which

is commonly applied in engineering.

14
IV. Numerical simulations

In this section, the control strategy of the TVC for upper stage with a gimbaled thruster

designed previously is simulated numerically.

A. Simulation Parameters

The parameters are mb  3600kg , mk  10kg , mm  30kg . Fm  5000N , Fg  25N .

rmz    1.107 0.0025 0.0022 m , rg  1m . rmc   0.1 0 0 m is the


T
rmx
T T
rmy

placement vector of the mass center for the GT expressed in frame f m . The inertia matrix of

upper stage ( I1 ), the gimbal ( I 2 ) and the GT ( I 3 ) are expressed as follows.

 4800 4 6.4  0.5 0 0 0.75 0 0


I1   4 6000 2.6  kg  m , I 2   0 0.7 0  kg  m , I 3   0 1.3 0  kg  m 2
  2   2 

 6.4 2.6 5800  0 0 0.6  0 0 1.4

The command direction of the thrust vector is  d d   0


T
0  ; The initial altitude of
T

the upper stage orbit is 1120km ; The initial attitude of the GT is     5
T
5  ; The
T

initial gimbal angles is     53.5174


T
0.1551  , The initial gimbal angular rate is
T

   0 0
T

T
/s .

The parameters of the designed controller are

k1  diag 1.5 1.5 1.5 2 2  , K  diag  0.25 0.25 0.25 0.25 0.25

ε  diag  0.0001 0.0001 0.0001 0.0001 0.0001

B. Simulation Results and Analysis

The simulation results are demonstrated in Fig. (4)~Fig. (8).

15
As a part of the thrust vector control, the simulation results of the special attitude control are

shown as follows. The attitude of the upper stage and its error in the desired attitude is shown in

Fig. 4, which illustrates that the upper stage’s attitude σ converges towards the desired attitude

σ d . It is shown that the attitude control of upper stage during orbit transfer is available. Fig. 5a)

implies the time history of the elevation angle and the azimuth angle of the thrust vector, and

Fig. 5b) verifies the angles converge towards the command angles determined by the guidance

system. The other part of thrust vector control is the gimbal control of the GT. The time histories

of the GT’s attitude and gimbal angles demonstrated in Fig. 6 indicate that the gimbal rotation of

the GT converges towards a stabilized state. Fig.4~Fig.6 reveals the thrust vector control of

upper stage is complete by the special attitude control of upper stage and the gimbal control of

the GT. Fig. 7 demonstrates the time histories of the attitude control torques acting on the upper

stage, showing that the torques about zb and yb axis are converge. For the torque about xb

provided by the RCS thrusters, the roll attitude approaches the corresponding limit loops, whose

radius is a near-zero constant. Fig. 8 shows the time histories of the gimbal rotation control

torques provided by electrical motors and indicates that the control torques are converge.

Fig.6~Fig.8 depicts the GT’s thrust vector passing through the mass center of the upper stage is

satisfied by the gimbal control of the GT. In addition to the definition of the special attitude

control process of the upper stage, the thrust vector of GT aligning with the command direction

is accomplished displays in Fig.5 when the special attitude control is achieved shown in Fig.4.

In summary, Fig. 4 to Fig.8 indicate that controlling the attitude of the upper stage and the

gimbal rotation of the GT, the thrust vector of the GT not only converges towards the command

direction determined by the guidance system, but also passes through the upper stage mass center,

avoiding the error of velocity vector and the disturbance torques.

16
60 10

Attitude error of upper stage/()




Attitude of upper stage/()


8
40  
6 

20  4

2
0
0

-20 -2
0 10 20 30 40 50 0 10 20 30 40 50
t/s t/s
a) b)

Fig. 4 Attitude (a) and attitude error (b) of upper stage

Azimuth and elevation error of GT/()


100 10
Azimuth and elevation of GT/()


80 
5 

60

40 0

20
-5
0

-20 -10
0 10 20 30 40 50 0 10 20 30 40 50
t/s t/s
a) b)

Fig. 5 Angles (a) and angles error (b) of GT’s thrust vector

8 20
 
Gimbal angles of GT/()

6  0 
Attitude of GT/()

4
-20
2

-40
0

-2 -60
0 10 20 30 40 50 0 20 40 60
t/s t/s
a) b)

17
Fig. 6 Attitude (a) and gimbal angles (b) of GT

Gimbal control torques of GT/(Nm)


Attitude control torques/(Nm) 200 3
T bx Tg
2
100 T by Tm
T bz 1

0 0

-1
-100
-2

-200 -3
0 10 20 30 40 50 0 10 20 30 40 50
t/s t/s
Fig. 7 Attitude control torques Fig. 8 Gimbal control torques

V. Conclusion

The disturbance torques, which is produced by the thrust vector misalignments and used to

control the attitude of the upper stage, is designed positively by controlling the gimbal angle. The

convergence of the attitude controller ensures the controller itself to eliminate the effects of the

thrust vector misalignments, and the desired attitude of the upper stage is designed to keep the

thrust vector direction of the GT same with the command’s after the stabilization of the

controller. The RCS thrusters in roll direction were employed to provide the control torque about

the roll axis. The gimbal rotations of the two-axis gimbal and the GT were controlled by the

electrical motor. A thrust vector control procedure including the upper stage attitude control and

the GT’s gimbal control was proposed. The control strategy guarantees that the thrust vector of

the GT not only aligns with the command direction but also passes through the upper stage’s

mass center. That is to say, the thrust vector of the upper stage accomplishes the goal to provide

the orbit transfer impetus effectively and do not produce disturbance torques. Numerical

simulations verified the effectiveness of the proposed control algorithm. The TVC of upper stage

presented in this paper will decrease the disturbance torques which act on the upper stage,

18
increase the accuracy of orbit transfer and decrease the consumption of fuel in the practical

launch mission.

The problem discussed in this paper was realized in recent launch missions. Our method and

results in this paper can be utilized in the TVC of the upper stage during orbit transfer in the

mission of One-vehicle with Multi-satellite in the future. This paper only concentrates on the

TVC of the upper stage. Detailed relations between the TVC and the orbit transfer control are of

great significance and worthy of further research.

Acknowledgments

As a part of One-Vehicle with Multi-Satellite Project, this work is supported by Shanghai

Institute of Aerospace Control Technology and National Natural Science Foundation (NO.

11272027).

References

1
A. Penchuk, S. Croopnick, “Digital Autopilot for Thrust Vector Control of the Shuttle

Orbital Maneuvering System,” Journal of Guidance, Control, and Dynamics, Vol. 6, NO. 6, 1983,

pp. 436-441.
2
D. N. Counter, B. C. Brinton, “Thrust Vector Control for the Space Shuttle Solid Rocket

Motor,” AIAA/SAE 11th Propulsion Conference, Anaheim, California, 1975, pp. 1-6.
3
Dale E. Schinstock, Douglas A. Scott, Tim A. Haskew, “Modeling and Estimation for

Electromechanical Thrust Vector Control of Rocket Engines,” Journal of Propulsion and Power,

Vol. 14, NO. 4, 1998, pp. 440-446.


4
Clinton B. F. Ensworth, “Thrust Vector Control for Nuclear Thermal Rockets,” 49th

AIAA/ASME/SAE/ASEE Joint Propulsion Conference & Exhibit, San Jose, CA, 2013, pp. 1-21.

19
5
Jon Millard, Brian Reed, “Implementation of the Orbital Maneuvering System Engine and

Thrust Vector Control for the European Service Module,” 50th AIAA/ASME/SAE/ASEE Joint

Propulsion Conference & Exhibit, Cleveland, OH, 2014, PP. 1-13.


6
Jaime Rubio Hervas, Mahmut Reyhanoglu, “Thrust Vector Control of a three-axis stabilized

upper-stage rocket with fuel slosh dynamics,” Acta Astronautica, Vol. 98, 2014, pp.120-127.
7
Bong Wie, “Thrust Vector Control Design for a Liquid Upper Stage Spacecraft,” Journal of

Guidance, Control, and Dynamics, Vol. 8, NO. 5, 1984, pp. 566-572.


8
Bong Wie, “Thrust Vector Control Analysis and Design for Solar-Sail Spacecraft,” Journal

of Spacecraft and Rocket, Vol. 44, NO. 3, 2007, pp. 545-557.


9
Petri K. Toivanen, Pekka janjunen, “Spin Plane Control and Thrust Vectoring of Electric

Solar Wind Sail,” Journal of Propulsion and power, Vol. 28, NO. 1, 2013, pp. 178-185.
10
Kougen Ma, Mehrdad N. Ghasemi-Nejhad, “Precision Positioning of a Parallel Manipulator

for Spacecraft Thrust Vector Control,” Journal of Guidance, Control, and Dynamics, Vol. 28,

NO. 1, 2005, pp. 185-188.


11
L. Felicetti, M. Sabatini, A. Pisculli, et al , “Adaptive Thrust Vector Control during On-

Orbit Servicing,” AIAA Space 2014 Conference and Exposition, San Diego, CA, 2014, pp. 1-18.
12
Michael Garrison, Mark Davis, Scott Steffan, “Human-Rated Upper Stage Thrust Vector

Control System Architecture Selection,” 43rd AIAA/ASME/SAE/ASEE Joint Propulsion

Conference & Exhibit, Cincinnati, OH, 2007, PP. 1-20.


13
Suchitra P, Priya C Kurian, Rini Jones S B, “Optimal Controller based Servo System Design

for the Thrust Vector Control of Liquid Propellant Engine of three Stage Launch Vehicle,”

International Conference on Magnetics, Machines & Drives, 2014, pp. 1-6

20
14
Tae Won Hwang, Chang-Su Park, Min-Jea Tahk, Hyochoong Bang, “Upper Stage Launch

Vehicle Servo Controller Design Considering Optimal Thruster Configuration,” AIAA

Guidance, Navigation, and Control Conference and Exhibit, Austin, Texas, 2003, pp. 1-8
15
Tobin C. Anthony, Bong Wie, “Pulse-Modulated Control Synthesis for a Flexible

Spacecraft,” Journal of Guidance, Control, and Dynamics, Vol. 13, NO. 6, 1989, pp. 1014-1022.
16
Zhaohui Wang, Lei Jin, Yinghong Jia. “Attitude Control Technology of Upper Stage during

Orbital Transfer,” China Space Science and Technology, NO. 5, 2014, PP. 1-9.
17
Gangbing Song, Nick V. Buck, Brij N. Agrawal, “Spacecraft Vibration Reduction Using

Pulse-Width and Pulse-Frequency Modulated Input Shaper,” Journal of Guidance, Control, and

Dynamics, Vol. 22, NO. 3, 1999, pp. 433-440.


18
Qinglei Hu, Guangfu Ma, “Vibration Suppression of Flexible Spacecraft During Attitude

Maneuvers,” Journal of Guidance, Control, and Dynamics, Vol. 28, NO. 2, 2005, pp. 377-380.

Highlights:

1) The rotary actuator is used due to the higher gimballing ability.

2) Coupled mathematical model of the upper stage as a multi-body system.

3) A control system consisting of the special attitude control and gimbal rotation.

4) The desired attitude computing method and the gimbal steering law.

21

You might also like