asen 5070 statistical orbit determination i fall 2012 professor jeffrey s. parker
DESCRIPTION
ASEN 5070 Statistical Orbit Determination I Fall 2012 Professor Jeffrey S. Parker Professor George H. Born Lecture 18: CKF, Numerical Compensations. Announcements. Homework 6 and 7 graded asap Homework 8 announced today. - PowerPoint PPT PresentationTRANSCRIPT
CCARColorado Center for
Astrodynamics Research
University of ColoradoBoulder 1
ASEN 5070Statistical Orbit Determination I
Fall 2012
Professor Jeffrey S. ParkerProfessor George H. Born
Lecture 18: CKF, Numerical Compensations
CCARColorado Center for
Astrodynamics Research
University of ColoradoBoulder 2
Homework 6 and 7 graded asap Homework 8 announced today.
Exam 2 on the horizon. Actually, it’s scheduled for 11/8! A week from Thursday.
We still have quite a bit of material to cover. Exam 2 will cover:
◦ Batch vs. CKF vs. EKF◦ Probability and statistics (good to keep this up!)◦ Observability◦ Numerical compensation techniques, such as the Joseph and Potter
formulation.◦ No calculators should be necessary◦ Open Book, Open Notes
Announcements
CCARColorado Center for
Astrodynamics Research
University of ColoradoBoulder 3
Quiz 13 Review
CCARColorado Center for
Astrodynamics Research
University of ColoradoBoulder 4
Quiz 13 Review
CCARColorado Center for
Astrodynamics Research
University of ColoradoBoulder 5
Quiz 13 Review
CCARColorado Center for
Astrodynamics Research
University of ColoradoBoulder 6
Quiz 13 Review
Only 46% of the class got this right
CCARColorado Center for
Astrodynamics Research
University of ColoradoBoulder 7
Quiz 13 Review
CCARColorado Center for
Astrodynamics Research
University of ColoradoBoulder 8
Quiz 13 Review
CCARColorado Center for
Astrodynamics Research
University of ColoradoBoulder 9
Quiz 13 Review
CCARColorado Center for
Astrodynamics Research
University of ColoradoBoulder 10
Quiz 13 Review
0% of the class got this right!
CCARColorado Center for
Astrodynamics Research
University of ColoradoBoulder 11
Quiz 13 Review
Observable quantities:• Semimajor axis of both orbits• Eccentricity of both orbits• True anomaly of both satellites• Inclination difference between the orbits• Node difference between the orbits• Arg Peri difference between the orbits
CCARColorado Center for
Astrodynamics Research
University of ColoradoBoulder 12
Quiz 14 Review
CCARColorado Center for
Astrodynamics Research
University of ColoradoBoulder 13
Quiz 14 Review
100% of the class got this right!
CCARColorado Center for
Astrodynamics Research
University of ColoradoBoulder 14
Quiz 14 Review
CCARColorado Center for
Astrodynamics Research
University of ColoradoBoulder 15
Quiz 14 Review
CCARColorado Center for
Astrodynamics Research
University of ColoradoBoulder 16
Quiz 14 Review
CCARColorado Center for
Astrodynamics Research
University of ColoradoBoulder 17
Quiz 14 Review
CCARColorado Center for
Astrodynamics Research
University of ColoradoBoulder 18
Quiz 14 Review
CCARColorado Center for
Astrodynamics Research
University of ColoradoBoulder 19
Quiz 14 Review
CCARColorado Center for
Astrodynamics Research
University of ColoradoBoulder 20
Quiz 14 Review
CCARColorado Center for
Astrodynamics Research
University of ColoradoBoulder 21
Quiz 14 Review
Probably the hardest question on this quiz and 78% got it right.
CCARColorado Center for
Astrodynamics Research
University of ColoradoBoulder 22
Homework 8
Due in 9 days
Assignment #8
CCARColorado Center for
Astrodynamics Research
University of ColoradoBoulder
CKF vs. Batch 1 1
, , , , ,k k k k kn n n p n p n nn nx P y K P
The Kalman (sequential) filter differs from the batch filter as follows:
1. It uses in place of . 2. It uses in place of
H 1,k kt t 0,kt t
Or if we don’t reinitialize, then
10 1 0, ,k kt t t t
1,k kt t i.e., we must invert at each stage
This means that is reinitialized at each observation time.
1 0,kt t
CCARColorado Center for
Astrodynamics Research
University of ColoradoBoulder
Processing Observations One at a TimeGiven a vector of observations at we may process these as scalar observations by performing the measurement update times and skipping the time update.
1p pp
kt
Algorithm1. Do the time update at kt
2. Measurement update
1 1 1, ,Tk k k k k kP t t P t t
1 1ˆ,k k k kx t t x
CCARColorado Center for
Astrodynamics Research
University of ColoradoBoulder
Processing Observations One at a Time
2a. Process the 1st element of . ComputekY *1 1 1 ,ky Y G
where
CCARColorado Center for
Astrodynamics Research
University of ColoradoBoulder
Processing Observations One at a Time
2 1
2 1
ˆHere k k
k k
x x
P P
2. Do not do a time update but do a measurement update by processing the 2nd element of (i.e. and are not mapped to )
kY 1ˆkx 1k
P1kt
Compute
etc.
Process pth element of . kY Compute
CCARColorado Center for
Astrodynamics Research
University of ColoradoBoulder
Processing Observations One at a Time
let ˆ ˆ ,pk kx x
pk kP P
Time update to and repeat this procedure
Note: must be a diagonal matrix. Why?
If not we would apply a “Whitening Transformation” as described in the next slides.
kR
1kt
CCARColorado Center for
Astrodynamics Research
University of ColoradoBoulder
Whitening Transformation
0E TE R
(1)Where is any sort of observation error covariance
R
CCARColorado Center for
Astrodynamics Research
University of ColoradoBoulder
Whitening Transformation
0E TE R
(1)
Factor where is the square root of TR SS S R
is chosen to be uppertriangularS
multiply Eq. (1) by 1S
Where is any sort of observation error covariance
R
CCARColorado Center for
Astrodynamics Research
University of ColoradoBoulder
Whitening Transformation
0E TE R
(1)
Factor where is the square root of TR SS S R
is chosen to be uppertriangularS
multiply Eq. (1) by 1S
let 1 ,y S y 1S
Where is any sort of observation error covariance
R
CCARColorado Center for
Astrodynamics Research
University of ColoradoBoulder
Whitening Transformation
Now 1 0E S E
1 1T T T TE S E S S RS
1 ,T TS SS S
Hence, the new observation has an error with zero mean and unit variance. We would now process the new observation and use .
1y S yy
CCARColorado Center for
Astrodynamics Research
University of ColoradoBoulder 32
Notice the decomposition in the whitening algorithm. ◦ S is a “square root” of R.
Several ways of computing S. A good way is the Cholesky Decomposition
(Section 5.2).
Major benefit of S vs. R: if the condition number of R is large, say 16, the condition number of S is half as large, say 8.
It’s much more accurate to invert S than R.
Cholesky DecompositionTR SS
CCARColorado Center for
Astrodynamics Research
University of ColoradoBoulder 33
Let M be a symmetric positive definite matrix, and let R be an upper triangular matrix computed such that
By inspection,
Cholesky Decomposition
Etc. See 5.2.1
CCARColorado Center for
Astrodynamics Research
University of ColoradoBoulder 34
M:
Use Matlab’s “chol” function to compute R:
Cholesky Decomposition
CCARColorado Center for
Astrodynamics Research
University of ColoradoBoulder 35
Consider the case where we have the least squares condition
If M is poorly conditioned, then inverting M will not produce good results.
If we decompose M first, then the accuracy of the solution will improve.
Let Then
Easy to forward-solve for z and then we can backward-solve for x-hat.
Cholesky Decomposition
CCARColorado Center for
Astrodynamics Research
University of ColoradoBoulder 36
Notice that the Cholesky Decomposition includes square roots. These are very expensive operations!
A square root free algorithm is given in 5.2.2.
Idea:
Cholesky Decomposition
CCARColorado Center for
Astrodynamics Research
University of ColoradoBoulder 37
Processing an observation vector one element at a time.
Whitening Cholesky
Quick Break
Next topic: Joseph, Potter, …
Questions?
CCARColorado Center for
Astrodynamics Research
University of ColoradoBoulder 38
Least squares estimation began with Gauss
1963: Kalman’s sequential approach◦ Introduced minimum variance◦ Introduced process noise◦ Permitted covariance analyses without data
Schmidt proposed a linearization method that would work for OD problems◦ Supposed that linearizing around the best estimate trajectory is better than linearizing
around the nominal trajectory
1970: Extended Kalman Filter
Gradually, researchers identified problems.◦ (a) Divergence due to the use of incorrect a priori statistics and unmodeled parameters.◦ (b) Divergence due to the presence of nonlinearities.◦ (c) Divergence due to the effects of computer round-off.
Kalman Filter History
CCARColorado Center for
Astrodynamics Research
University of ColoradoBoulder 39
Numerical issues cause the covariance matrix to lose their symmetry and nonnegativity
Possible corrections:◦ (a) Compute only the upper (or lower) triangular entries and force symmetry◦ (b) Compute the entire matrix and then average the upper and lower fields◦ (c) Periodically test and reset the matrix◦ (d) Replace the optimal Kalman measurement update by other expressions
(Joseph, Potter, etc)◦ (e) Use larger process noise and measurement noise covariances.
Kalman Filter History
CCARColorado Center for
Astrodynamics Research
University of ColoradoBoulder 40
Potter is credited with introducing square root factorization.◦ Worked for the Apollo missions!
1968: Andrews extended Potter’s algorithms to include process noise and correlated measurements.
1965 – 1969: Development of the Householder transformation◦ Worked for Mariner 9 in 1971!
1969: Dyer-McReynolds filter added additional process noise effects.◦ Worked for Mariner 10 in 1973 for Venus and Mercury!
Kalman Filter History
CCARColorado Center for
Astrodynamics Research
University of ColoradoBoulder 41
Batch:◦ Large matrix inversion w/ potential poorly-conditioned matrix.
Sequential:◦ Lots and lots of matrix inversions w/ potentially poorly-conditioned
matrices.
EKF:◦ Divergence: If the observation data noise is too large.◦ Saturation: If covariance gets too small, new data stops influencing the
solution.
Numerical Issues
CCARColorado Center for
Astrodynamics Research
University of ColoradoBoulder 42
Replace
with
This formulation will always retain a symmetric matrix, but it may still lose positive definiteness.
Joseph Formulation
CCARColorado Center for
Astrodynamics Research
University of ColoradoBoulder
Equivalence of Joseph and Conventional Formulation of the Measurement Update of P
Need to show that
TKH P KH KRK KH P
First derive the Joseph formulation
x̂ x K y Hx subst. y Hx
x̂ x K Hx Hx
x̂ x KH x x K
subtract from both sides and rearrangex
x̂ x x x KH x x K
KH x x K
CCARColorado Center for
Astrodynamics Research
University of ColoradoBoulder
form ˆ ˆ TP E x x x x
T TP KH P KH KRK
where
,TP E x x x x TR E
0,TE x x since is independent of ix t it
Equivalence of Joseph and Conventional Formulation of the Measurement Update of P
CCARColorado Center for
Astrodynamics Research
University of ColoradoBoulder
Next, show that this is the same as the conventional Kalman
T T T T TKH P KH KRK KH P KH PH K KRK
T T T T TKH P PH K KHPH K KRK
T T T TKH P PH K K HPH R K
Equivalence of Joseph and Conventional Formulation of the Measurement Update of P
CCARColorado Center for
Astrodynamics Research
University of ColoradoBoulder
Next, show that this is the same as the conventional Kalman
T T T T TKH P KH KRK KH P KH PH K KRK
T T T T TKH P PH K KHPH K KRK
T T T TKH P PH K K HPH R K
subst. for 1T TK PH HPH R
1T T T T T TKH P PH K PH HPH R HPH R K
T T T TKH P PH K PH K
KH P
Equivalence of Joseph and Conventional Formulation of the Measurement Update of P
CCARColorado Center for
Astrodynamics Research
University of ColoradoBoulder 47
Question on Joseph?
Next: Demo!
CCARColorado Center for
Astrodynamics Research
University of ColoradoBoulder
Example Illustrating Numerical Instability of Sequential (Kalman) Filter (see 4.7.1)The following example problem from Bierman (1977) illustrates the numerical characteristics of several algorithms we have studied to date. Consider the problem of estimating and from scalar measurements and 1x 2x 1z 2z
1 1 2 1z x x
2 1 2 2z x x
Where and are uncorrelated zero mean random variables with unit variance. If they did not have the above traits we could perform a whitening transformation so that they do. In matrix form
1 2
1 1 1
2 2 2
14.7.20
1 1z xz x
CCARColorado Center for
Astrodynamics Research
University of ColoradoBoulder
The a priori covariance associated with our knowledge of is assumed to be
1
2
xX
x
2 20
1 00 1
P I
Where and . The quantity is assumed to be
small enough such that computer round-off produces
1 0 1
21 1 4.7.21 This estimation problem is well posed. The observation provides an estimate of which, when coupled with should accurately determine . However, when the various data processing algorithms are applied several diverse results are obtained.
1z2z1x 2x
Example Illustrating Numerical Instability of Sequential (Kalman) Filter (see 4.7.1)
CCARColorado Center for
Astrodynamics Research
University of ColoradoBoulder
Let the gain and estimation error covariance associated with be denoted as and respectively. Similarly the measurement is associated with and respectively. Note that this is a linear parameter (constant) estimation problem.
Hence,
1z 1k1P 2z 2k 2P
, kt t I
1k kP P
1k k k kP I k h P 1
1 1 1 1,2T Tk k k k k kk P h h P h k
We will process the observation one at a time. Hence,
Example Illustrating Numerical Instability of Sequential (Kalman) Filter (see 4.7.1)
CCARColorado Center for
Astrodynamics Research
University of ColoradoBoulder
Note: Since we will process the observations one at a time, the matrix inversion in previous equation is a scalar inversion. The exact solution is
1
1 0 1 1 0 1T Tk P h h P h
1
2 21 11 1
1 2
11 4.7.221 2
k
where 2 2 1
Example Illustrating Numerical Instability of Sequential (Kalman) Filter (see 4.7.1)
CCARColorado Center for
Astrodynamics Research
University of ColoradoBoulder
1 2
21 4.7.231
P
where 21 2
Example Illustrating Numerical Instability of Sequential (Kalman) Filter (see 4.7.1)
The estimation error covariance associated with processing the first data point is
21 2
11 11 2
P I I
1 1 1 0P I k h P
CCARColorado Center for
Astrodynamics Research
University of ColoradoBoulder
2 2 2 1P I k h P
The exact solution for is given by:2P
2
2
1 2 11 (4.7.24)1 2
Example Illustrating Numerical Instability of Sequential (Kalman) Filter (see 4.7.1)Processing the second observation:
1
2 1 2 2 1 2 1T Tk Ph h Ph
where 2 21 2 2 2
2
1 211
CCARColorado Center for
Astrodynamics Research
University of ColoradoBoulder
The conventional Kalman filter yields (using )21 1
1
1k
21
11P I I
1 2
0(4.7.25)P
Example Illustrating Numerical Instability of Sequential (Kalman) Filter (see 4.7.1)
Note that is no longer positive definite. The diagonals of a matrix must be
positive. However does exist since .1P PD
11P
1 0P
CCARColorado Center for
Astrodynamics Research
University of ColoradoBoulder
1
2 1 2 2 1 2 1T Tk Ph h Ph
1
11 2
2 2 2 1P I k h P 1 11 (4.7.26)1 11 2
Compute P2 using the Conventional Kalman
Now is not positive definite ( ) nor does it have positive terms on the diagonal. In fact, the conventional Kalman filter has failed for these observations.
2P 2 0P
Example Illustrating Numerical Instability of Sequential (Kalman) Filter (see 4.7.1)
CCARColorado Center for
Astrodynamics Research
University of ColoradoBoulder
The Joseph formulation yields
2 2 2 1 2 2 2 2T TP I k h P I k h k k
Example Illustrating Numerical Instability of Sequential (Kalman) Filter (see 4.7.1)
CCARColorado Center for
Astrodynamics Research
University of ColoradoBoulder
The Batch Processor yields
21 11 1 1
1I
12
2
2 1
1 2 1
12
1
101
ii
Ti PhhP
Example Illustrating Numerical Instability of Sequential (Kalman) Filter (see 4.7.1)
P2 for the batch
2
1 2 (1 3 )1 3 2 4
P
CCARColorado Center for
Astrodynamics Research
University of ColoradoBoulder
P2 for the batch
2
1 2 (1 3 )1 3 2 4
P
To order , the exact solution for is2P
2
1 1 1 2 1 31 2
1 2 1 3 2 4P
Example Illustrating Numerical Instability of Sequential (Kalman) Filter (see 4.7.1)
CCARColorado Center for
Astrodynamics Research
University of ColoradoBoulder
Summary of Results2P
Exact to order 1 2 (1 3 )
(1 3 ) 2 4
Conventional Kalman
1 111 11 2
Joseph
1 2 (1 3 )(1 3 ) 2
Batch
1 2 (1 3 )(1 3 ) 2 4
Example Illustrating Numerical Instability of Sequential (Kalman) Filter (see 4.7.1)
CCARColorado Center for
Astrodynamics Research
University of ColoradoBoulder 60
Homework 6 and 7 graded asap Homework 8 announced today.
Exam 2 on the horizon. Actually, it’s scheduled for 11/8! A week from Thursday.
We still have quite a bit of material to cover. Exam 2 will cover:
◦ Batch vs. CKF vs. EKF◦ Probability and statistics (good to keep this up!)◦ Observability◦ Numerical compensation techniques, such as the Joseph and Potter
formulation.◦ No calculators should be necessary◦ Open Book, Open Notes
The End