Planeamento de Missões Espaciais

Assignment 1: Lambert’s problem - 2D

Instituto Superior Técnico, MSc Aerospace Engineering, 2014/2015
Prof: P. J. S. Gil

Ricardo Filipe Pereira Gomes 69385

October 29, 2014

Abstract: A fundamental problem in astrodynamics is concerned with the de-

termination of an orbit, having a specified flight time and connecting two position
vectors. The present assignment describes how can we solve this problem, which is
frequently referred to as Lambert’s problem

1 Problem Definition

For interplanetary direct missions a

minimum-energy transfer, the Hohmann
transfer orbit, should be considered.
However this interplanetary trajectory
could be obviously to much long for a
viable mission. This creates the need of
finding orbits and corresponding veloci-
ties for different geometries, which could
be done by using, for instance, so-called Figure 1: Geometry of two-dimensional
Lambert targeting. Moreover, this is an interplanetary trajectory (not to scale)[3]
efficient way to get 1st order solution of
an orbit transfer from planet A to planet 1.1 Algorithm
B, assuming a kepler orbit (central body,
no perturbations, no thrusting). In Lambert problem at least two solutions
may exist; one could go clockwise and the
other anti-clockwise. For both directions
In the Lambert’s problem, the initial of motion an orbit, that satisfies the time
position, final position, and the desired requirement, can be found. Since all the
time for the transfer between the two po- planets orbit around the Sun in an anti-
sitions is known. Solving Lambert’s prob- clockwise direction, it is wise to choose
lem allows the calculation of the iner- the same direction for the motion of the
tial velocities at the initial and final po- satellite. Otherwise, by choosing clock-
sitions, Fig 1, as well as the value for the wise direction it would require an enor-
semi-major axis. Such parameters will be mous amount of energy to compensate
used to define the orbital elements of the planet’s velocity. In the subsequent al-
desired transfer orbit. Consequently, it gorithm this could be chosen by select-
has important applications in the areas ing a positive(anti-clockwise) or negative
of rendezvous, targeting, guidance, and (clockwise) value for ∆θ
preliminary orbit determination. [1] The literature is able to provide many

1
forms for an algorithm to solve the Lam- However E1 and E2 are unknown.
bert’s Problem. It started with the clas- Hence, one solution for this problem is
sical method devised by Gauss, attractive to normalize ∆t and a
because of its simplicity. Unfortunately it r
has two major flaws: the method is sin- 8µ
T = ∆t (1.3)
gular for a transfer angle of 180 deg and s3
the convergence rate is extremely slow,
s
when that angle is not very small. There- x2 = a − (1.4)
2a
fore [Battin and Vaughan,1983] have de-
vised an algorithm that takes away these Where,
flaws [2]. Then again, [Gooding,1990] did r1 + r2 + c
this as well, albeit in a different way, also s= (1.5)
2
providing the possibility of implementing
multiple revolutions. Although, to sim- So, equation 1.2 can be now expressed
plify, in this assignment the algorithm in the non-dimensional variables T and x
will only consider a maximum of one sin- 2(x − qz − d/y)
gle revolution. For these reasons, the T = (1.6)
Elam
algorithm developed by [Gooding, 1990]
will be applied and summarized in the In equation 1.6 the new parameters
rest of this section [3]. are all dependent of x, except q, which
As a starting point to find the Lam- follows directly from a geometry problem.
bert’s problem solution, let’s consider a √
r1 r2
generic transfer orbit, Fig 2 q= cos(∆θ/2) (1.7)
s

Elam = x2 − 1 = Elam (x) (1.8)

p
y= Elam = y(x) (1.9)

q
z= 1 − q 2 + q 2 x2 = z(x) (1.10)

Determine d involves computationally

more effort. First one must calculate the
Figure 2: Transfer orbit geometry based following new parameters,
in Lambert’s problem [4]

where, f = y(z − qx) = f (x) (1.11)

g = xz − qElam = g(x) (1.12)

c2 = r12 + r22 − 2r1 r2 cos(∆θ) (1.1)
Then if Elam < 0(and thus an elliptic
Thus the time of flight and the orbital orbit)
elements are related by [5] d = arctan(f /g) (1.13)
s and if Elam > 0 (and thus an hyper-
a3 bolic orbit)
∆t = (E2 − E1 − e(sin(E2 − sinE1 ))
µ
(1.2) d = ln(f + g) (1.14)

2
 In the implemented code the single ar-
gument arctangent function will be sub-
stitute by atan2 which uses two argu-
ments instead of one to gather informa-
tion on the signs of the inputs in order
to return the appropriate quadrant of the
computed angle. Thus, the following ex-
pression derived from the tangent half-
angle formula can be used to define atan2.

p
f 2 + g2 − g
atan2(f, g) = 2arctan
f
(1.15)

In spite of do not knowing the value

of x, I know the value that T should have, Figure 3: Plot of T vs x for selected val-
which will lead to, ues of θ/π and correspondent q and m as
indicated [6])

To simplify the problem only satellites

8µ 2(x − qz − d/y) that perform a single revolution will be be
T (= 3 ∆t) = ⇒
s Elam consider, i.e., m = 0. Moreover, to im-
2(x − q(x)z(x) − d(x)/y(x)) prove the algorithm by making it faster
F (x) = T −
Elam (x) and with less iterations, [Gooding,1990]
=0 developed a strategy to determine x0 ,
(1.16) when m = 0.
By considering,
x − qz − d/y
T0 = 2 , where x = 0
Elam
By considering Formula 1.16 and us- (1.18)
ing the Newton-Rapson method the value
of x can be easily computed,1 Thus,
T0 − T
If T0 ≥ T ⇒ x0 = T0 ,
4T
r (1.19)
8µ
with T = ∆t
F (xi ) s
xi+1 = xi − 0 (1.17)
F (xi )
T − T0
If T0 < T ⇒ x0 = − ,
T − T0 + 4
r
8µ
with T = ∆t
Though the starting value, x0 , has to s
(1.20)
be available before iteration can begin.[6]
To ensure even faster convergence appro- After convergence of x, the value for
priate starters have been developed by the semi-major axis of the transfer orbit
[Gooding,1990], as shown in Figure 3 can now be determined.
1
To compute the derivative of F 0 (x) it was used the Matlab function (diff)

3
 s/2
a= (1.21) If a < 0 (Hyperbola) ⇒
1 − x2
Then, with some more calculation 1 − ra1
E1 = acosh( ) and (1.32)
other relevant elements to characterize e
the conic can be found [4]. With the 1 − ra2
E2 = acosh( )
aid of the three new parameters presented e
bellow, the inertial velocities can be com-
In conclusion, by using the former al-
puted, accordingly to Figure 1
gorithm, the conic trajectory described
q by the satellite can be determined know-
γ= µs/2 (1.22)
ing only the two position vectors of the
planets and the time of flight. The Mat-
ρ = (r1 − r2 )/c (1.23) lab program to solve Lambert’s Problem
is in Section 4.1

q ∆θ
σ = 2 r1 r2 /c2 sin( ) (1.24) 2 Testing and Validation
2
Thus, The convergence of x is one of the re-
quirements that should be taken into ac-
count in order to validate the algorithm
V1,rad = γ[(qz −x)−ρ(qz +x)]/r1 (1.25) described in Section 1.1

V2,rad = −γ[(qz − x) + ρ(qz + x)]/r2 By using Newton-Rapson iteration

(1.26) method with an initial value for x0 given
V1,tan = γσ(z + qx)/r1 (1.27) by the [Gooding 1990] starters, Eq. 1.18
to Eq. 1.20, it is observed that the fi-
V2,tan = γσ(z + qx)/r2 (1.28)
nal solution converge very quickly, for a
Since the inertial velocities and the very small relative error of the semi-major
specific relative angular momentum are axis, a (Fig 4).
known, it is easy to determine the eccen-
tricity of the transfer orbit [5].

h = V1,tan r1 = V2,tan r2 (1.29)

s
h2
e= 1− (1.30)
µa
Other important parameter to charac-
terize the orbit is the eccentric anomaly,
which can be now calculated by using the
previous data [5].

If a > 0 (Ellipse) ⇒
Figure 4: Relative error of the semi-major
1 − ra1
E1 = acos(
e
) and (1.31) axis in % as function of the number of
iterations, considering r1 = 1AU, r2 =
1 − ra2
E2 = acos( ) 1.5AU, ∆θ = 45 deg, ∆t = 90days
e

4
 Moreover, to validate the results ob- tion. Richmond, Virginia, USA,
tained by this algorithm a comparative William Bell Inc., 1998.
analysis between the Hohmann trans-
fer trajectory and the Lambert prob- [2] Richard H.Battin and Robin
lem should be performed. Considering M. Vaughan, An elegant
an Hohmann transfer between Earth Lambert Algorithm. Cam-
and Mars,2 where, r1 = 1AU, r2 = bridge,Massachusetts, USA,
1.5AU, ∆θ = 180 deg, ∆t = 225.231days Journal of Guidance, Control,
as input, the return values of the algo- and Dynamics 1984 7:6 , 662-670.
rithm used are,
[3] Jeroen Melman, Trajectory Opti-
[ a (AU) , e , E1 [ deg ] , E2 [ deg ] ] = mization for a Mission to Nep-
=1.250 ,0.200 ,0.000 ,180.000 tune and Triton - Thesis Re-
Lastly, as intended, the conic ob- port. Delft, The Netherlands, TU
tained corresponds exactly to the ellip- Delft, 18 January 2007.
tic orbit of an Hohmann transfer between
Earth and Mars, which can be considered [4] R. Noomen, Space Mission De-
as a proof of the validity of this algorithm. sign: Lambert Targeting AE4-878
V4.8. Delft, The Netherlands, TU
Delft, 2 December 2013.
3 Conclusion
In conclusion this algorithm is a good [5] Wakker K. F, Astrodynamics I
way to determine the conic that takes - Lecture notes AE4-874. Delft,
a specific time for the satellite to trans- The Netherlands, TU Delft - Fac-
fer from one planet to other. The main ulty of Aerospace Engineering,
limitation of this algorithm is that it is 2007a.
based in a two-body problem approxima-
tion. However results are good enough to [6] R.H. Gooding, A procedure for
be considered for preliminary orbit deter- the solution of Lambert’s orbital
mination. Future developments to this boundary-value problem. Farnbor-
algorithm could take into consideration ough, Hants, England Royal
the relative inclination between the two Aerospace Establishment, 9 Jan-
planets and multiple revolutions. uary 1990.

The estimated time to realize this as- [7] James F. Jordan, The application
signment was about 25 hours. of Lambert’s Theorem to the So-
lution of Interplanetary Transfer
References Problem - Technical report No.32-
521. Pasadena, California, Jet
[1] J.M.A. Danby, Fundamental of Propulsion Laboratory, 1 Febru-
Celestial Mechanics 2nd edi- ary 1964.

2
Earth and Mars orbit are assumed to be circular

5
 4 Appendix
4.1 Matlab Algorithm
4.1.1 main.m

1 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2 % LAMBERT PROBLEM SOLVER
3 % Author : R i c a r d o F i l i p e P e r e i r a Gomes
4 % S t u de n t Number : 69385
5 % Assignment 1 : main f u n c t i o n
6 % F i l e name : main .m
7 % Version 1.0
8 % 27 October 2014
9 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
10
11
12 %D e f a u l f v a l u e s :
13 r 1= 1 ; %[AU]
14 r 2= 1 . 5 ; %[AU]
15 t h e t a= 1 8 0 ; % [ deg ]
16 miu= 1 3 2 7 1 2 4 4 0 0 1 8 ; %[ km^ 3 . s ^ −2]
17 d_t= 2 5 5 . 2 3 1 ; %[ days ]
18
19
20 [ a , e , E1 , E2 ] = l a m b e r t ( r1 , r2 , t h e t a , miu , d_t )
21 f p r i n t f ( ’ [ a (AU) , e , E1 [ deg ] , E2 [ deg ] ] = % 3 . 3 f , % 3 . 3 f , % 3 . 3 f , % 3 . 3 f \n ’ , a , e , E1 , E2 ) ;

4.1.2 lambert.m

1 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2 % LAMBERT PROBLEM SOLVER
3 % Author : R i c a r d o F i l i p e P e r e i r a Gomes
4 % S t u de n t Number : 69385
5 % Assignment 1 : S o l v e Lambert ’ s problem 2D: g i v e n two p l a n e t s ,
6 % d e t e r m i n e a c o n i c t h a t t a k e s a s p e c i f i c time t o
7 % transfer
8 % F i l e name : l a m b e r t .m
9 % Version 1.0
10 % 27 October 2014
11 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
12
13 f u n c t i o n [ a_AU, e , E1 , E2 ] = l a m b e r t ( r1 , r2 , t h e t a , miu , d_t )
14 %%%%%%%%%%%%%%%%%%%%%%% INPUTS %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
15 % r 1 − d i s t a n c e t o p l a n e t 1 i n [AU]
16 % r 2 − d i s t a n c e t o p l a n e t 2 i n [AU]
17 % t h e t a − a r c between a r r i v a l and d e p a r t u r e i n [ deg ]
18 % miu − g r a v i t a c i o n a l p a r a m e t e r o f t h e c e n t r a l body i n [ km^ 3 . s ^ −2]
19 % d_t i s t h e time t r a n s f e r i n days
20
21 %%%%%%%%%%%%%%%%%%%%%% OUTPUTS %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
22 % a_UA − semi−major a x i s o f t h e t r a n s f e r o r b i t [AU]
23 % e − e c c e n t r i c i t y of the t r a n s f e r o r b i t
24 % E1 − e c c e n t r i c anomaly on d e p a r t u r e [ deg ]
25 % E2 − e c c e n t r i c anomaly i n a r r i v a l [ deg ]
26 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
27
28 t = cputime ; %a l l o w s t o know t h e time o f computation
29
30 % Convert v a r i a b l e s :
31 AU=149597871000; %Value o f one a s t r o n o m i c a l u n i t i n [m]
32 deg_r=p i / 1 8 0 ;
33 r_deg =180/ p i ;
34
35 r 1=r 1 ∗AU; %c o n v e r t i n t o [m]
36 r 2=r 2 ∗AU; %c o n v e r t i n t o [m]
37 t h e t a=t h e t a ∗ deg_r ; %c o n v e r t i n t o rad
38 miu = miu ∗ 1 e9 ; % c o n v e r t i n t o [m^ 3 . s ^ −2]
39 d_t=d_t ∗ 8 6 4 0 0 ; %Convert days i n t o s e c o n d s

6
 40
41 %%%%%%%%%% Begin o f t h e a l g o r i t h m %%%%%%%%%%%%%%%%%%%%%%%%%
42
43 c=s q r t ( r 1 ^2+ r 2 ^2−2∗ r 1 ∗ r 2 ∗ c o s ( t h e t a ) ) ;
44 s =( r 1+r 2+c ) / 2 ;
45 T=s q r t ( 8 ∗ miu/ s ^ 3 ) ∗d_t ;
46
47 q=s q r t ( r 1 ∗ r 2 ) / s ∗ c o s ( t h e t a / 2 ) ; %km
48
49 % To compute t h e i n i t i a l v a l u e o f x0 i n o r d e r t o o p t i m i z e t h e
50 % a l g o t i t h m s i n c e t h e i n i t i a l v a l u e w i l l came c l o s e r t o t h e f i n a l
51 % s o l u t i o n , then f a s t e r c o n v e r g e n c e
52
53 x0 =0;
54 Elam0=x0 ^2 −1;
55 y0=s q r t ( abs ( Elam0 ) ) ;
56 z0=s q r t (1−q^2+(q∗ x0 ) ^ 2 ) ;
57 f 0=y0 ∗ ( z0−q∗ x0 ) ;
58 g0=x0 ∗ z0−q∗Elam0 ;
59 i f ( Elam0<0) %E l l i p t i c o r b i t
60 d0=atan2 ( f 0 , g0 ) ; % atan2 ( f , g )=atan ( f / g ) The p u r p o s e o f u s i n g two
61 % arguments i n s t e a d o f one i s t o g a t h e r i n f o r m a t i o n on
62 % the s i g n s of the inputs in order to return the
63 % a p p r o p r i a t e quadrant o f t h e computed a n g l e , which i s
64 % not p o s s i b l e f o r t h e s i n g l e −argument a r c t a n g e n t f u n c t i o n
65 else
66 d0=l o g ( f 0+g0 ) ; %H y p e r b o l i c o r b i t
67 end
68
69 %I n i t i a l v a l u e o f x
70 T0=2∗(x0−q∗ z0−d0 / y0 ) /Elam0 ;
71
72 i f ( T0>T)
73 x0=T0 ∗ ( T0−T) / ( 4 ∗T) ; %x0
74 else
75 x0=−(T−T0 ) / (T−T0+4) ; %x0
76 end
77
78 %D e f i n i t i o n o f t h e s i m b o l i c v a r i a b l e s b e f o r e i t e r a t e
79 syms x ;
80 Elam=x ^2 −1;
81 y=s q r t ( abs ( Elam ) ) ;
82 z=s q r t (1−q^2+(q∗x ) ^ 2 ) ;
83 f=y ∗ ( z−q∗x ) ;
84 g=x∗ z−q∗Elam ;
85
86 %C a l c u l a t e i n e r t i a l velocities
87
88 gamma=s q r t ( miu∗ s / 2 ) ;
89 rho =(r1−r 2 ) / c ;
90 sigma=2∗ s q r t ( r 1 ∗ r 2 / c ^ 2 ) ∗ s i n ( t h e t a / 2 ) ;
91
92 % I t i s used m a t l a b F u n c t i o n t o Convert s y m b o l i c e x p r e s s i o n
93 % to f u n c t i o n handle
94
95 V1_rad=gamma ∗ ( ( q∗ z−x )−rho ∗ ( q∗ z+x ) ) / r 1 ; %r a d i a l v e l o c i t i y d e p a r t u r e
96 V1_rad=m a t l a b F u n c t i o n ( V1_rad ) ;
97 V1_tan=gamma∗ sigma ∗ ( z+q∗x ) / r 1 ; %t a n g e n c i a l v e l o c i t y d e p a r t u r e
98 V1_tan=m a t l a b F u n c t i o n ( V1_tan ) ;
99
100 V2_rad=−gamma ∗ ( ( q∗ z−x )+rho ∗ ( q∗ z+x ) ) / r 1 ; %r a d i a l v e l o c i t y a r r i v a l
101 V2_rad=m a t l a b F u n c t i o n ( V2_rad ) ;
102 V2_tan=gamma∗ sigma ∗ ( z+q∗x ) / r 2 ; %t a n g e n c i a l v e l o c i t y a r r i v a l
103 V2_tan=m a t l a b F u n c t i o n ( V2_tan ) ;
104
105
106 % With syms atan2 d o e s n t work , need t o be decomposed .
107 % For t h i s i t w i l l be used t h e t a n g e n t h a l f −a n g l e f o r m u l a
108 % ATAN2(Y,X) = 2∗ atan ( ( s q r t (X^2 + Y^ 2 ) − X) /Y) ;
109 % This must be used f o r a c o m p l e t c o v e r a g e between 0 and 2 p i

7
110
111 d_e = 2∗ atan ( ( s q r t ( f ^2 + g ^ 2 ) − g ) / f ) ; %i f eliptic orbit
112 d_h=l o g ( f+g ) ; %i f h y p e r b o l i c o r b i t
113
114 % D e f i n i t i o n o f x u s i n g Newton−Rapson method
115 % x ( i +1)=x ( i )−F( x ) /F ’ ( X)
116
117 %I f t h e o r b i t i s e l l i p t i c
118 F_e=T−(2∗( x−q∗ z−d_e/y ) /Elam ) ;
119 dF_e= d i f f ( F_e ) ;
120 aux1=F_e/dF_e ;
121 aux1=m a t l a b F u n c t i o n ( aux1 ) ;
122
123 %I f t h e o r b i t i s h y p e r b o l i c
124 F_h=T−(2∗( x−q∗ z−d_h/y ) /Elam ) ;
125 dF_h= d i f f (F_h) ;
126 aux2=F_h/dF_h ;
127 aux2=m a t l a b F u n c t i o n ( aux2 ) ;
128
129
130 Elam=m a t l a b F u n c t i o n ( Elam ) ;
131 e r r o r=I n f ;
132 i =1;
133 aux ( i )=x0 ;
134 i n t e r ( i ) =1;
135
136 w h i l e e r r o r >1e −15 % Does not compromise computation time and i m p l i e s
137 % a v e r y s m a l l e r r o r f o r t h e semi−major a x i s
138 % ( s m a l l e r than a few k i l o m e t e r s )
139
140 i f ( Elam ( aux ( i ) ) <0)
141 aux ( i +1)=aux ( i )− aux1 ( aux ( i ) ) ; %E l l i p t i c O r b i t
142 else
143 aux ( i +1)=aux ( i )− aux2 ( aux ( i ) ) ; %H y p e r b o l i c o r b i t
144 end
145
146 e r r o r=abs ( aux ( i +1)−aux ( i ) ) ; %Compute i t e r a t i o n e r r o r
147
148 % Creates the v a r i a b l e s f o r f i g u r e 4 in the assignment
149 % i n t e r ( i +1)=i +1;
150 % e r r o ( i )=abs ( s /2∗1/(1 − aux ( i +1) ^ 2 ) . . .
151 % −s /2∗1/(1 − aux ( i ) ^ 2 ) ) / ( s /2∗1/(1 − aux ( i +1) ^ 2 ) ) ∗100
152
153 i=i +1;
154 end
155
156 % e r r o ( i )=abs ( s /2∗1/(1 − aux ( i ) ^ 2 ) . . .
157 % −s /2∗1/(1 − aux ( i −1) ^ 2 ) ) / ( s /2∗1/(1 − aux ( i ) ^ 2 ) ) ∗100
158 x=aux ( i ) ;
159 a= s /2∗1/(1 − x ^ 2 ) ; %Semi−major a x i s a f t e r c o n v e r g e n c e o f x
160 a_AU=a /AU; %Convert i n t o AU
161
162 V1_rad=V1_rad ( x ) ; %R a d i a l d e p a r t u r e v e l o c i t y a f t e r c o n v e r g e n c e o f x
163 V1_tan=V1_tan ( x ) ; %T a n g e n c i a l d e p a r t u r e v e l o c i t y a f t e r c o n v e r g e n c e o f x
164 V2_rad=V2_rad ( x ) ; %R a d i a l a r r i v a l v e l o c i t y a f t e r c o n v e r g e n c e o f x
165 V2_tan=V2_tan ( x ) ; %T a n g e n c i a l a r r i v a l v e l o c i t y a f t e r c o n v e r g e n c e o f x
166
167 h=V1_tan∗ r 1 ; % R e l a t i v e a n g u l a r momentum
168 e=s q r t (1 −(h ^ 2 ) / ( miu∗ a ) ) %e c c e n t r i c i t y o f t h e t r a n s f e r o r b i t
169
170 % C a l c u l u s o f t h e e c c e n t i c anomaly
171 i f a>0 % I f e l l i p s e
172 E1=a c o s ((1 − r 1 / a ) / e ) ∗ r_deg ;
173 E2=a c o s ((1 − r 2 / a ) / e ) ∗ r_deg ;
174 e l s e %I f h y p e r b o l a
175 E1=a c o s h ((1 − r 1 / a ) / e ) ∗ r_deg ;
176 E2=a c o s h ((1 − r 2 / a ) / e ) ∗ r_deg ;
177 end
178
179 %%Used t o c a l c u l a t e t h e f l i g h t path a n g l e on d e p a r t u r e and a r r i v a l p o i n t

8
180 %f p a 1=atan2 ( V1_rad , V1_tan ) ∗ r_deg ;
181 %f p a 2=atan2 ( V2_rad , V2_tan ) ∗ r_deg ;
182
183
184 % To a n a l y s e t h e r a t e o f c o n v e r g e n c e
185 % s e m i l o g y ( i n t e r , e r r o , ’ bo ’ ) ;
186 % x l a b e l ( ’ Number o f i t e r a t i o n s ’ ) % x−a x i s l a b e l
187 % y l a b e l ( ’ R e l a t i v e e r r o r f o r t h e semi−major a x i s [ % ] ’ )
188
189
190
191
192 time = cputime−t ; %h e l p f u l l t o a n a l y s e computation time