Massachusetts Institute of Technology Instrumentation Laboratory Cambridge, Massachusetts

Transcription

1 Massachusetts Institute of Technology Instrumentation Laboratory Cambridge, Massachusetts Space Guidance Analysis Memo #-64 TO: SGA Distribution FROM: Gerald M. Levine DATE: February, 964 SUBJECT: The Transition Matrix for a Circular Orbit The transition matrix is a means of determining the propagation of small deviations from a reference orbit. Thus, 6 (6v where or and Sv are the position and velocity deviations, Sr initial deviations, and 43is the 6 X 6 transition matrix. The 'matrix satisfies the matrix differential equation and 6v the I / where I and are the 3 X 3 identity and ero matrices, respectively, and G is the gravity gradient matrix. The initial value of the 4. matrix is the 6 X 6 identity matrix. Let the position and velocity vectors have components in an inertial axis system as follows: - -

2 r= Let the reference trajectory be a circular orbit in a central force field with initial conditions r ) r = ( Then the reference orbit is given by ( cos wt EREF = r sin wt where w is the orbital angular rate. The gravity vector is given by g - r 3 r where p is the gravitational constant and r is the position of the spacecraft and is given by r = r REF + 6 the G matrix is then -2-

3 - 3 cos t cot -3 sin cot cos cot G =-w 2-3 sin cot cos cot - 3 sing cot (2) where 2 -. r 3 Since the propagation of errors in the xy plane is independent of the errors along the axis, the 'matrix may be written as follows: xx 5 xy (hxk g6xr yx YY YX OyST I:5 Z Z Cb;x ky Yx (kik X 3-, Sr* S, Sr YY 95i Let xx (hxy Oxi C5xy yx YY 3 3, ;. 95. ky is,- yx yy yx * / ( 3 ) -3-

4 and = (4) Then, using Eqs. (2), (3), and (4), Eq. () may be written as the following two equations: xy -( 2 ( - 3 cost w t) 3co 2 sin cot cos wt 3co 2 sin cot cos cot -w 2 ( - 3 sing wt) (5) = -co 2 / (6) The initial values of I y and t are the identity matrices of the appropriate dimensions. Differentiating Eq. (6) yields = -co 2 4> which has solution cos wt sin cot -w sin wt cos wt /

5 To solve Eq, (5), let a column of the matrixbe xy w / I Then, each column satisfies the same differential equation as the t xy matrix. Thus, p ( 4 w -w 2 { ( - 3 cost wt) p - (3 sin cot cos cot) q] -w 2 [ -(3 sin wt cos wt) p + ( - 3 sin 2 wt) (7) Eliminating the variables s and w from Eq. (7) yields = - w 2 ( - 3 cos t wt) p - (3 sin wt cos wt) (8) q = 2 :-( 3 sin wt cos wt) p+ (- 3 sin 2 wt) q (9) Let = wt and denote differentiation with respect to 8 by a prime. Then, Eqs. (8) and (9) may be written as follows: It p - 3 ( + 3 cos 2 ) p - ( - 2_ 2. 2 ) q = () q. - ( 3 sm 2) p - ( - 3 cos 2) q =

6 variables Equations () and () may be written in terms of the complex u = p + iq v= p - Multiplying Eq.() by i, and forming the sum and difference of the result with Eq. () gives " 3 2 u - u - e v= 2 2 (2) tt v - v- 3 e -2i u = 2 2 (3) After some algebraic manipulation of Eqs. (2), (3) and the first two derivatives of Eq. (2), the following equation for u alone is obtained Int It, II I u - 4i u i u = (4) If a solution u= e i k is assumed for u and substituted into Eq. (4), the following fourth degree polynomial for k results: 2 k 4 4k3 + 5k - 2k = or k(k - I) (k - 2) = The general solution of Eq. (4) is then given by -6-

7 u = (A + i B) + (A + i B) e + (A2 + i B2) (A3 + i B3 ) e 2i9 (5) where the A's and B's are real constants. Since there are only four independent constants, the relationships between the eight A's and B's are obtained by substituting Eq. (5) and its conjugate into Eq. (2). Equation (5) then becomes u = (A + i B ) + (A + i B ) e - 3 Aḻ O o l e i 2 (A - i B) e 2i (6) Expressions for p(t) and q(t) are obtained by taking real and imaginary parts of Eq. (6). Differentiation of p(t) and q(t) yields s(t) and w(t). The general solution of Eq. (7) is then A( - cos 2) +Ai (cos + -- sin ) - -3'- Bo sin 2 - Bi sin A O sin 2 + A (sin - cos ) + Bo (+ cos 2 ) + B cos.) I A sin Ai (sin + 3 cos ) - 2 Bo cos 2 - B cos,, Ao cos 2-2- A.(cos 2-3 sin ) - Bo sin 26) - sine; (7) The initial value of is obtained by setting t equal to ero in Eq. (7). Thus, AA + Al. 3 u ± B O + B 3 - co ( 4 Bo + B) -6.(- 2 n + w A) 3 u -7-

8 Each to has one component equal to unity and three ero components, allowing the four constants to be evaluated. The resulting solution of Eq. (5) as follows: + 2C(-C)+ 3S (9--s): S(-C) Lis(2-7 [-+c(2-c)+3s(b-s)] = xy 25(-C)-3C(9-S) :-C(-C) 2t [-C(2-C)] a-js(2-c)-3c(-s)] co[ S(+C)+3C(-S)] ' co[l+c(- 2C)] +2C(-C) SO. - C)+3 -S) (8) w2-c(+c)+3s(-s) :w[s(-2c)] 2S(-C) i -C(-C)+3S(-S) Where S = sin C = cos It is of interest to express Eq. (8) in nominal local vertical coordinates, i. e., in an axis system rotating about the axis at a rate equal to w. To accomplish this, each of the 2 X 2 submatrices of 4 y must be premultiplied by the following matrix: cos sin -sin cos e Then, the transition matrix in the local vertical system+ is given by x y -cos wt sinwt. sm t 2 ( - cos wt) 2 sin cot - 3 wt -( - 2 cos cot) - ( - cos wt) (4 sin wt - 3 wt) xy -w(sinwt - 3 wt) w ( - cos wt) 2 - cos cot -(2 sinwt - 3 wt) (- cos wt) -w sin wt - sin w t - ( - 2 cos wt) -8-