Conference PaperPDF Available

Recursive nonlinear set-theoretic estimation based on pseudo ellipsoids

Abstract

The problem of estimating a vector x_ of unknown quantities based on a set of measurements depending nonlinearly on x_ is considered. The measurements are assumed to be taken sequentially and are corrupted by unknown but bounded uncertainties. For this uncertainty model, a systematic design approach is introduced, which yields closed-form expressions for the desired nonlinear estimates. The estimates are recursively calculated and provide solution sets χ containing the feasible sets, i.e., the sets of all x_ consistent with all the measurements available and their associated bounds. The sets χ are tight upper bounds for the exact feasible sets and are in general not convex and not connected. The proposed design approach is versatile and the resulting nonlinear filter algorithms are both easy to implement and efficient.
Recursive Nonlinear Set–Theoretic Estimation
Based on Pseudo Ellipsoids
Uwe D. Hanebeck
Institute of Automatic Control Engineering
Technische Universit¨at M¨unchen
80290 unchen, Germany
Uwe.Hanebeck@ieee.org
Abstract
In this paper, the problem of estimating a vector
x
of unknown quantities based on a set of measure-
ments depending nonlinearly on x
is considered. The
measurements are assumed to be taken sequentially
and are corrupted by unknown but bounded uncer-
tainties. For this uncertainty model, a systematic de-
sign approach is introduced, which yields closed–form
expressions for the desired nonlinear estimates. The
estimates are recursively calculated and provide solu-
tion sets X containing the feasible sets, i.e., the sets
of all x
consistent with all the measurements available
and their associated bounds. The sets X are tight
upper bounds for the exact feasible sets and are in
general not convex and not connected. The proposed
design approach is versatile and the resulting nonlin-
ear filter algorithms are both easy to implement and
efficient.
1 Introduction
Estimating the state of a dynamic system based on
a sequence of uncertain measurements is a standard
problem in many applications. Usually, this problem
is approached in a stochastic setting. Alternatively,
set–theoretic methods can be used by assuming a pri-
ori bounds on the uncertainties. Estimation then con-
sists of constructing sets of possible states, which are
consistent with the a priori bounds and the measure-
ments. Good overviews about this topic can be found
in [1, 3, 19].
Most work has been done in set–theoretic state es-
timation for linear systems [17, 18, 21, 22]. Appli-
cations in the field of speech processing are found in
[4, 5]. Robotic applications are discussed in [7, 16]. A
comparison of stochastic and set–theoretic estimation
is given in [6].
In the case of nonlinear systems, the complex sets
resulting from the estimation procedure are either
approximated by simple–shaped sets, e.g. ellipsoids,
boxes [15, 12], polytopes [20], or by the union of simple
sets [11, 12].
In [20], an approach similar to the Extended
Kalman Filter (EKF) is pursued. As in the EKF,
the nonlinear state equations are linearized about the
current state estimate. Unlike the EKF, linearization
errors are not neglected, but rather considered as ad-
ditional exogeneous disturbances. Estimation is per-
formed recursively and provides polytopes as approx-
imation of the posterior feasible set.
The procedure in [15] works without linearization
of the nonlinear state equations and provides (recur-
sively) the smallest axis–aligned box enclosing the
posterior feasible set.
In [11], the posterior feasible set is characterized by
enclosing it between internal and external unions of
boxes on the basis of interval analysis. Recursive es-
timation has not been addressed. A more advanced
version of this approach with lower computational
complexity has been introduced in [12].
In this paper, a new nonlinear filtering algorithm
for nonlinear systems is proposed, that does not rely
in any way on linearization. In addition, the new ap-
proach is not based on a grid or on propagating par-
ticles, but provides a finite–dimensional closed–form
representation of the resulting complex–shaped sets.
This includes nonconvex sets or sets that are not even
connected. When applying the new filter recursively
to a sequential stream of measurements, the size of the
analytical representation of the resulting sets does not
grow with the number of measurements.
The key idea of the proposed filter is to trans-
form the original N –dimensional space S to an L
dimensional hyperspace S
with L>N.Thisre-
sults in an N–dimensional manifold U
, called the
universal manifold, in the L–dimensional transformed
space S
. Complex–shaped subsets of the original
N–dimensional space are then represented by N
dimensional submanifolds of U
in the space S
.
These submanifolds are defined by the intersection
of L–dimensional simple–shaped sets, e.g. ellipsoids,
with the universal manifold U
. Furthermore, the
nonlinear measurement equation is transformed to a
linear one in the hyperspace S
. Hence, nonlinear fil-
tering can be performed by a linear filter operating in
the transformed space S
.
Section 2 formulates the problem of nonlinear set–
theoretic estimation. In Sec. 3 the concept of modeling
complex–shaped sets is introduced. The nonlinear fil-
tering algorithm is given in Sec. 4. The performance of
the new nonlinear filtering algorithm is demonstrated
by estimating the parameters of a SCARA–type robot
manipulator in Sec. 5.
2 Problem Formulation
Consider a nonlinear discrete–time dynamic system
with system state x
k
(not directly observable) at time
step k.Measurementsˆz
k
of the system output are
taken at time instants k =1, 2,... according to the
nonlinear measurement equation
ˆz
k
= h
k
(x
k
)+v
k
(1)
with measurement uncertainty v
k
, which represents
exogenous noise sources or model parameter uncer-
tainties.
The uncertainties v
k
, k =1, 2,..., are assumed to
be bounded by a known set V
k
according to v
k
∈V
k
.
The set can be of complicated shape, i.e., can be
nonconvex or not connected.
The goal is to estimate at each time instant k the
state x
k
based on all available measurements ˆz
l
for
l =1, 2,... ,k. Of course, a recursive estimation pro-
cedure is preferred, which calculates a state estimate
based on the estimate at the previous time step and
the current measurement. Therefore, it is not required
to store and reprocess all measurements. Further-
more, instead of trying to construct point estimates,
we prefer to calculate at each time instant k all states
that are compatible with the measurements and their
corresponding uncertainties.
On a theoretical level, the problem can easily be
solved: Let X
s
k1
denote the set of all states com-
patible with all the measurements up to time step
k 1 and their respective uncertainties. Furthermore,
X
m
k
denotes the set of states defined solely by the
measurement at time k according to
X
m
k
= {x
k
z
k
h
k
(x
k
) ∈V
k
} .
Then the estimate X
s
k
is given by the intersection
X
s
k
= X
s
k1
∩X
m
k
.
However, representing these sets in practical applica-
tions at least approximately by a finite set of parame-
ters is not a trivial task. On one hand, the parameter
set should not be too large, even more, the approx-
imation should degrade gracefully with a decreasing
number of parameters. On the other hand, the num-
ber of parameters should not be permanently growing
with an increasing number of incoming measurements.
Hence, the remainder of this paper is concerned with
a new parametric representation of complex–shaped
sets and an efficient procedure for calculating the cor-
responding parameters.
3 Pseudo Ellipsoids
The key idea of this paper is to represent an un-
certainty X
k
with a complicated shape in the N
dimensional original space S by a simpler shaped un-
certainty X
k
in an L–dimensional hyperspace S
with
L>N.Pointsx
k
in S are related to points x
k
in S
via a nonlinear transformation T (.) according to
x
k
= T (x
k
)=
T
1
(x),T
2
(x), ... , T
L
(x)
T
.
Hence, T
(.) defines an N –dimensional manifold U
in
an L–dimensional space.
In addition, L–dimensional sets X
k
of simple shape
are defined in the transformed space S
. Here, ellip-
soidal sets according to
X
k
=
x
k
:(x
k
ˆx
k
)
T
(X
k
)
1
(x
k
ˆx
k
) 1
are used, where ˆx
k
is the ellipsoid midpoint and X
k
is
a symmetric positive definite matrix.
The intersection of an ellipsoid X
k
with the mani-
fold U
defines a submanifold of U
, which, in turn,
defines a complicated set in the original space S.
Remark 3.1 A complex–shaped set in the original
space S is defined by both the transformation T
(.) and
the pseudo ellipsoid X
k
.
In many cases, the type of transformation T
(.)re-
sults directly from the nonlinearities considered. For
example, when considering polynomial nonlinearities,
a polynomial transformation is used. For trigonomet-
ric nonlinearities, a trigonometric transformation can
be used.
However, to simplify application of the new filtering
approach, a generic transformation is helpful. For that
purpose, Bernstein polynomials are used, since their
approximation capabilities are sufficient for a large
class of nonlinearities. In addition, they lead to better
conditioned calculations than standard polynomials.
Multidimensional Bernstein polynomials are defined
on the basis of one–dimensional Bernstein polynomi-
als, which on the interval [l, r]aregivenby
H
n
i
(x)=
n
i

l x
l r
i
r x
r l
ni
for i =0,... ,n.With
x
k
=
x
k,1
x
k,2
... x
k,N
T
,
the above transformation is defined by
T
i
(x
k
)=
N
j=1
H
L
j
1
i
j
(x
k,j
) ,
for i
j
=0,... ,L
j
1, j =1,... ,N, L =
N
j=1
L
j
,
and i =
N
j=1
i
j
.
4 Filtering
Based on the concept of pseudo ellipsoids, which
represent complex–shaped sets in the original space S
by pseudo ellipsoids in the hyperspace S
, the nonlin-
ear filter step can now be performed by a linear filter
in the hyperspace S
. For that purpose, a pseudo
linear expansion of the nonlinear measurement equa-
tion h
k
(.) is performed according to
h
k
(x
k
)=H
k
x
k
+ e
h
k
H
k
x
k
,
where e
h
k
represents the approximation error defined
by e
h
k
= h
k
(x
k
) H
k
x
k
.
In general, the expansion can be enhanced by an
additional nonlinear transformation g
(.)ofthemea-
surements according to
g
z
k
v
k
)=g(h
k
(x
k
)) .
The left hand side can be approximated by
g
z
k
v
k
)=ˆz
k
G
k
v
k
e
v,
k
ˆz
k
G
k
v
k
,
where ˆz
k
and G
k
are nonlinear functions of ˆz
k
and
v
k
is a nonlinear function of v
k
. e
v,
k
accounts for the
approximation error.
The right hand side is again approximated accord-
ing to
g
(h
k
(x
k
)) = H
k
x
k
+ e
h,
k
H
k
x
k
with approximation error e
h,
k
.Asaresult,themea-
surement equation in the hyperspace is obtained ac-
cording to
z
k
= H
k
x
k
+ e
h,
k
+ G
k
v
k
+ e
v,
k

w
k
with w
k
representing the total uncertainty.
Let the set of all predicted states be given by the
set X
p
k
, which is defined in the transformed space S
by
X
p,
k
=
x
k
:(x
k
ˆx
p,
k
)
T
(E
p,
k
)
1
(x
k
ˆx
p,
k
) 1
.
Furthermore, let w
k
be bounded by the set
W
k
=
w
k
:(w
k
)
T
(W
k
)
1
w
k
1
.
Then, the set defined by the measurement is given by
X
m,
k
=
x
k
:(ˆz
k
H
k
x
k
)
T
(W
k
)
1
z
k
H
k
x
k
) 1
.
The fusion result is given by a set X
s,
k
(again an el-
lipsoid in the transformed space, but a set of compli-
cated shape in the original space!) that contains the
intersection of the ellipsoids X
p,
k
and X
m,
k
. Hence,
X
s,
k
is obtained by a linear set–theoretic filter in the
hyperspace S
[19]
X
s,
k
=
x
k
:(x
k
ˆx
s,
k
)
T
(E
s,
k
)
1
(x
k
ˆx
s,
k
) 1
with
ˆx
s,
k
x
p,
k
+ λ
k
E
p,
k
(H
k
)
T
W
k
+ λ
k
H
k
E
p,
k
(H
k
)
T
1
ˆz
k
H
k
ˆx
p,
k
,
and
E
s,
k
= d
k
P
s,
k
P
s,
k
= E
p,
k
λ
k
E
p,
k
(H
k
)
T
W
k
+ λ
k
H
k
E
p,
k
(H
k
)
T
1
H
k
E
p,
k
,
where
d
k
=1 + λ
k
λ
k
ˆz
k
H
k
ˆx
p,
k
T
W
k
+ λ
k
H
k
E
p,
k
(H
k
)
T
1
ˆz
k
H
k
ˆx
p,
k
.
Using this form of bounding ellipsoid for the exact
intersection of X
p,
k
, X
m,
k
in the transformed space S
offers the advantage that the resulting set X
s
k
(λ
k
) in
the original space S possesses the following property,
which is desirable in applications in the sense, that no
new uncertainty is introduced:
l
1
l
2
x
EE
x
Figure 1: Schematical top view of the considered
type of 2D SCARA robot manipulator with segment
lengths l
1
, l
2
and joint angles φ
1
, φ
2
.
Lemma 4.1 If the two sets X
p
k
und X
m
k
overlap, the
filtering result X
s
k
(λ
k
) contains the exact intersection
X
p
k
∩X
m
k
and is itself contained in their union X
p
k
X
m
k
. Hence, it holds
(X
p
k
∩X
m
k
) ⊂X
s
k
(λ
k
) (X
p
k
∪X
m
k
)
for all λ
k
[0, ].
The fusion parameter λ
k
is selected in such a way,
that a certain measure of the size of the set X
s
k
is min-
imized. (How to obtain the minimum volume ellipsoid
in a linear setting is discussed in [2].)
5 Simulation Example
We consider a SCARA–type robot manipulator
with two degrees of freedom according to Fig. 1. The
segment lengths l
1
, l
2
are assumed to be unknown and
are not amenable to direct measurements. Only uncer-
tain measurements of the distance r of the end–effector
from the origin are available for different angles φ
1
, φ
2
.
The position of the end–effector with respect to the
origin is given by
x
EE
=
l
1
cos(φ
1
)+l
2
cos(φ
1
+ φ
2
)
l
1
sin(φ
1
)+l
2
sin(φ
1
+ φ
2
)
.
Hence, the distance r of the end–effector from the
origin is given by
r =
l
2
1
+ l
2
2
+2l
1
l
2
cos(φ
2
) .
Given this nonlinear relation, the segment lengths l
1
,
l
2
are estimated (N = 2) based on measured distances
ˆr
k
, k =1, 2,... for different angles φ
2,k
.Themea-
surement uncertainties are assumed to be bounded
according to
|r
k
ˆr
k
|≤R,
which gives
ˆr
k
=
l
2
1
+ l
2
2
+2l
1
l
2
cos(φ
2,k
)+v
k
with v
2
k
R
2
.
The parameters used during the simulation are
l
1
= 500 mm ,
l
2
= 350 mm ,
R =40mm .
The filtering procedure starts with an axis–aligned
uncertainty box
X
p
1
=[0, 1000] × [0, 1000] mm
2
.
Sequentially, measurements ˆr
k
, k =1,... ,4, for
φ
2,1
=90
o
, φ
2,2
= 120
o
, φ
2,3
= 160
o
, φ
2,4
= 170
o
are used to update the initial estimate.
The function g
(.) has been selected to g(x)=
x
2
x
4
x
6
T
, T
i
(x), i =1,... ,L, are chosen as
multidimensional Bernstein polynomials. The results
are visualized
1
for n = 6 in Fig. 2, where n is the
order of the one–dimensional Bernstein polynomials,
which gives L = 49. Obviously, the result of the pro-
posed new filter is a tight approximation of the exact
estimation result at every time step.
6 Conclusions
In this paper, an efficient algorithm for the recur-
sive calculation of tight closed–form approximations of
the feasible sets in nonlinear set–theoretic estimation
problems has been presented. The resulting sets are
a much better approximation compared to simple sets
like hyper–rectangles or ellipsoids and are in general
of complex shape, i.e., nonconvex and not connected.
The same methodology has been applied to stochas-
tic nonlinear systems [10]. For the case of mixed
stochastic and set–theoretic uncertainties, the filter-
ing algorithms presented in [8, 9] will be generalized
to nonlinear systems.
References
[1] F. L. Chernousko, State Estimation for Dynamic
Systems, CRC Press, 1994.
[2] M.-F. Cheung, S. Yurkovich, K. M. Passino, “An
Optimal Volume Ellipsoid Algorithm for Parame-
ter Set Identification”, IEEE Tr ansactions on Au-
tomatic Control, Vol. 38, No. 8, pp. 1292–1296,
1993.
1
The result of updating the estimate based on measurement
k is denoted X
s
k
. However, in the next time step, the same
estimate is denoted X
p
k+1
to be consistent with the notation
introduced above.
Figure 2: Results of sequential estimation of the lengths l
1
, l
2
of the 2D SCARA robot manipulator.
[3] P. L. Combettes, “The Foundations of Set The-
oretic Estimation”, Proceedings of the IEEE,
Vol. 81, No. 2, pp. 182–208, 1993.
[4] J. R. Deller, Set Membership Identification in
Digital Signal Processing”, IEEE ASSP Maga-
zine, Vol. 6, pp. 4–20, 1989.
[5] J. R. Deller, M. Nayeri, S. F. Odeh, “Least–
Squares Identification with Error Bounds for
Real–Time Signal Processing and Control”, Pro-
ceedings of the IEEE, Vol. 81, No. 6, pp. 815–849,
1993.
[6] G.D.Hager,S.P.Engelson,S.Atiya,“OnCom-
paring Statistical and Set–Based Methods in Sen-
sor Data Fusion”, Proceedings of the 1993 IEEE
International Conference on Robotics and Au-
tomation, Atlanta, Georgia, pp. 352–358, 1993.
[7] U. D. Hanebeck, G. Schmidt, “Set–theoretic Lo-
calization of Fast Mobile Robots Using an An-
gle Measurement Technique”, Proceedings of the
1996 IEEE International Conference on Robotics
and Automation (ICRA’96), Minneapolis, Min-
nesota, pp. 1387–1394, 1996.
[8] U. D. Hanebeck, J. Horn, and G. Schmidt, “On
Combining Statistical and Set–Theoretic Estima-
tion”, Automatica, Vol. 35, No. 6, pp. 1101–1109,
1999.
[9] U. D. Hanebeck and J. Horn, “Fusing Information
Simultaneously Corrupted by Uncertainties with
Known Bounds and Random Noise with Known
Distribution”, Information Fusion, Vol. 1, No. 1,
pp. 55–63, 2000.
[10] U. D. Hanebeck, “Optimal Filtering for Polyno-
mial Measurement Nonlinearities with Additive
Non–Gaussian Noise”, Proceedings of the Amer-
ican Control Conference (ACC’2001), Arlington,
Virginia, 2001.
[11] L. Jaulin and E. Walter, “Set Inversion via In-
terval Analysis for Nonlinear Bounded–error Es-
timation”, Automatica, Vol. 29, No. 4, pp. 1053–
1064, 1993.
[12] L. Jaulin, “Interval constraint propagation with
application to bounded–error estimation”, Auto-
matica, Vol. 36, No. 10, pp. 1547–1552, 2000.
[13] T. H. Kerr, “Real–Time Failure Detection: A
Static Non–Linear Optimization Problem That
Yields a Two Elliposid Overlap Test”, Jour-
nal on Optimization Theory Applications,Vol.2,
pp. 509–536, 1977.
[14] M. Milanese, G. Belaforte, “Estimation Theory
and Uncertainty Intervals Evaluation in the Pres-
ence of Unknown but Bounded Errors: Linear
Families of Models and Estimates”, IEEE Trans-
actions on Automatic Control, Vol. 27, pp. 408–
414, 1982.
[15] M. Milanese, A. Vicino, Estimation Theory for
Nonlinear Models and Set Membership Uncer-
tainty”, Automatica, Vol. 27, No. 2, pp. 403–408,
1991.
[16] A. Preciado, D. Meizel, A. Segovia, M. Rombaut,
“Fusion of Multi–Sensor Data: A Geometric Ap-
proach”, Proceedings of the 1991 IEEE Interna-
tional Conference on Robotics and Automation,
Sacramento, California, pp. 2806–2811, 1991.
[17] A. K. Rao, Y.–F. Huang, S. Dasgupta, “ARMA
Parameter Estimation Using a Novel Recursive
Estimation Algorithm with Selective Updating”,
IEEE Transactions on Signal Processing, Vol. 38,
No. 3, pp. 447–457, 1990.
[18] A. Sabater, F. Thomas, Set Membership Ap-
proach to the Propagation of Uncertain Geomet-
ric Information”, Proceedings of the 1991 IEEE
International Conference on Robotics and Au-
tomation, Sacramento, California, pp. 2718–2723,
1991.
[19]
F. C. Schweppe, Uncertain Dynamic Systems.
Prentice–Hall, 1973.
[20] J. S. Shamma, K.–Y. Tu, “Approximate Set–
Valued Observers for Nonlinear Systems”, IEEE
Transactions on Automatic Contr ol, Vol. 42,
No. 5, pp. 648–658, 1997.
[21] A. Vicino, G. Zappa, “Sequential Approxima-
tion of Feasible Parameter Sets for Identification
with Set Membership Uncertainty”, IEEE Trans-
actions on Automatic Control, Vol. 41, No. 6,
pp. 774–785, 1995.
[22] E. Walter, H. Piet-Lahanier, “Exact Recursive
Polyhedral Description of the Feasible Parame-
ter Set for Bounded Error Models”, IEEE Trans-
actions on Automatic Control, Vol. 34, No. 8,
pp. 911–915, 1989.
... X 0,i to be a decreasing sequence with limit the optimal estimate X * 0 . Nonetheless, the settheoretic algorithm in its general form accounts for some difficulties: The performance of the above method depends on the initial guess X 0 , as well as on our knowledge regarding the sets D and E. Subsequently, representing the sets X 0 , D, E, R D,u i and C E,y i+1 (D E,y i+1 ) in practical applications -at least approximately-by a finite set of parameters, is not a trivial task [27]. Finally, the method does not provide any accuracy about the belief degree regarding the state estimates. ...
... Lastly, by substituting the Lie derivative terms (Appendix 7.1), equation (27) results in ...
Preprint
This paper conveys attitude and rate estimation without rate sensors by performing a critical comparison, validated by extensive simulations. The two dominant approaches to facilitate attitude estimation are based on stochastic and set-membership reasoning. The first one mostly utilizes the commonly known Gaussian-approximate filters, namely the EKF and UKF. Although more conservative, the latter seems to be more promising as it considers the inherent geometric characteristics of the underline compact state space and accounts -- from first principles -- for large model errors. We address the set-theoretic approach from a control point of view, and we show that it can overcome reported deficiencies of the Bayesian architectures related to this problem, leading to coordinate-free optimal filters. Lastly, as an example, we derive a modified predictive filter on the tangent bundle of the special orthogonal group $\mathbb{TSO}(3)$.
... Nonetheless, the set-theoretic algorithm in its general form accounts for some difficulties: The performance of the above method depends on the initial guesŝ  0 , as well as on our knowledge regarding the sets  and . Subsequently, representing the sets  0 , , , ℛ ,u i and ,y i+1 ( ,y i+1 ) in practical applications -at least approximately-by a finite set of parameters, is not a trivial task [47]. Finally, the method does not provide any accuracy about the belief degree regarding the state estimates. ...
Article
Full-text available
This paper conveys attitude and rate estimation without rate sensors by performing a critical comparison, validated by extensive simulations. The two dominant approaches to facilitate attitude estimation are based on stochastic and set-membership reasoning. The first one mostly utilizes the commonly known Gaussian approximate filters, namely the EKF and UKF. Although more conservative, the latter seems to be more promising as it considers the inherent geometric characteristics of the underline compact state space and accounts -from first principles- for large model errors. We address the set-theoretic approach from a control point of view, and we show that it can overcome reported deficiencies of the Bayesian architectures related to this problem, leading to coordinate-free optimal filters. Lastly, as an example, we derive a modified predictive filter on the tangent bundle of the special orthogonal group TSO(3).
... Second, stochastic and set-theoretic estimation approaches (Hanebeck [2001]) are applicable to quantify the estimation quality. In both cases, prerequisites are the quantification of the uncertainties of the dynamical system model as well as the measurement process. ...
Article
Full-text available
In this paper, techniques based on Carleman linearization are presented which are applicable to the design of controllers and state estimators for nonlinear dynamical systems. Originally, Carleman linearization has been developed to transform sets of polynomial ordinary differential equations into infinite dimensional linear system representations. After choosing suitable finite orders for truncation of Carleman linearization, finite dimensional linear and bilinear system representations are obtained for which the design of controllers and state estimators can be performed in a straightforward way. To visualize the practical applicability of the presented procedures, state as well as disturbance estimation for a linear axis driven by pneumatic muscle actuators is studied.
... In [Yang 2009] the noises are assumed to be confined to specified ellipsoidal sets. We can find similar approaches in [Garulli 1999, Hanebeck 2001. Non-linear system is more difficult. ...
Thesis
Full-text available
Modeling dynamic systems requires to account for uncertainties arising from noises impacting the measures and/or ! the dynamics, from lack of knowledge about disturbances, and also from uncertainties on parameter values (tolerance specifications, wear processes). Some of these uncertainties, like measurement noises, can be properly modelled in statistical terms but others are better characterized by bounds, without any additional feature. In this thesis, motivated by the above considerations, we consider the problem of integrating both statistical ans bounded uncertainties for discrete time linear systems. Building on the Interval Kalman Filter (IKF) developed by [Chen 1997], we propose significant improvements based on recent techniques of constraint propagation and set inversion which, unlike the IKF algorithm, allow us to obtain guaranteed results while controlling the pessimism of interval analysis. The improved filter is named iIKF. The iIKF filter has the same recursive structure as the classical Kalman filter and delivers an enclosure of all the possible optimal estimates and the covariance matrices. The previous IKF algorithm avoids the interval matrix inversion problem and consequently looses possible solutions. For the iIKF, we propose an original guaranteed method for the interval matrix inversion problem that couples the SIVIA (Set Inversion via Interval Analysis) algorithm and a set of constraint propagation problems. In addition, several mechanisms based on constraint propagation are implemented to limit the overestimation effect of interval propagation within the filter recursive structure. A fault detection algorithm based on the iIKF is proposed. It implements a semi-closed loop strategy which stops feeding the filter with observation corrupted by the fault as soon as it is detected. Through various examples, the advantages of the iIKF filter are presentedand the effectiveness of the of the fault detection algorithm is demonstrated.
... Second, stochastic and set-theoretic estimation approaches (Hanebeck [2001]) are applicable to quantify the estimation quality. In both cases, prerequisites are the quantification of the uncertainties of the dynamical system model as well as the measurement process. ...
Conference Paper
Full-text available
In this paper, techniques based on Carleman linearization are presented which are applicable to the design of controllers and state estimators for nonlinear dynamical systems. Originally, Carleman linearization has been developed to transform sets of polynomial ordinary differential equations into infinite dimensional linear system representations. After choosing suitable finite orders for truncation of Carleman linearization, finite dimensional linear and bilinear system representations are obtained for which the design of controllers and state estimators can be performed in a straight forward way. To visualize the practical applicability of the presented procedures, state as well as disturbance estimation for a linear axis driven by pneumatic muscle actuators is studied.
... Since the measurement equations are nonlinear and the uncertainties are large, a standard filtering approach based on linearization of the measurement equation cannot be applied. Instead, a new nonlinear set-theoretic estimation technique based on pseudo ellipsoids [8] is used. ...
Article
Full-text available
Within the existing GSM standard, several measurements are available that can be used for estimating the position of a cellular phone. First, the timing advance (TA) gives an estimate for the distance to the serving base station. Second, the signal strengths (RXLEV) of neighbouring base stations can also be interpreted as distance information. Both TA and RXLEV are subject to measurement errors caused for example by shadowing, reflections, and fast fading. Thus, a nonlinear set-theoretic estimation technique based on pseudo ellipsoids is applied. The uncertainty regions in the original space defined by the measurements are transformed into a hyperspace of higher dimension and described by pseudo ellipsoids. An approximation of the set intersection of the pseudo ellipsoids can be calculated recursively by a linear set-theoretic filter. The resulting pseudo ellipsoid is transformed back into the original space, and the position estimate is calculated as center of gravity of the resulting uncertainty region. The algorithm is evaluated based on the data of an extensive field trial in a rural area. Compared to pure cell ID, the accuracy is significantly increased by using TA and RXLEV, reducing the mean error by half.
... Hence, proven linear filtering techniques can be employed for solving the nonlinear estimation problem. A similar approach has been proposed for the case of nonlinear set–theoretic estimation in the case of unknown–but–bounded noise descriptions (Hanebeck, 2001b) and evaluated extensively in applications (Horn et al., 2002; Briechle and Hanebeck, 2003). ...
Article
We consider the problem of estimating the state of a discrete–time dynamic system comprising a linear system equation and a nonlinear measurement equation based on measurements corrupted by non–Gaussian noise. The problem is solved by recursively calculating the complete posterior density of the state given the measurements. For representing the resulting non–Gaussian posterior, a new exponential type density, the so called pseudo Gaussian density, is introduced. By converting the original nonlinear system to an equivalent linear representation in a higher–dimensional space, the parameters of the pseudo Gaussian posterior are obtained by means of a linear estimator operating in the higher–dimensional space. The resulting filtering algorithms are easy to implement and always guarantee valid posterior densities.
Article
State estimation techniques for centralized, distributed, and decentralized systems are studied. An easy-to-implement state estimation concept is introduced that generalizes and combines basic principles of Kalman filter theory and ellipsoidal calculus. By means of this method, stochastic and set-membership uncertainties can be taken into consideration simultaneously. Different solutions for implementing these estimation algorithms in distributed networked systems are presented. © 2014 Karlsruher Institut fur Technologie (KIT). All rights reserved.
Conference Paper
Full-text available
The theoretical and practical considerations of two common sensor data fusion methodologies (set based and statistically based parameter estimation) are compared. Their convergence behavior for a variety of simulated problems is examined. Robot localization systems implemented using both methods are described, and their performance is compared. It is concluded that set-based methods have performance that sometimes exceeds that of statistical methods, although this result is highly problem-dependent. These problem dependencies are characterized
Article
Full-text available
Real-time failure detection for systems having linear stochastic dynamical truth models is posed in terms of two confidence region sheaths. One confidence region sheath is about the expected no-failure trajectory; the other is about the Kalman estimate. If these two confidence regions of ellipsoidal cross section are disjoint at any time instant, a failure is declared. A test for two-ellipsoid overlap is developed which involves finding a single pointx* whose presence in both ellipsoids is necessary and sufficient for overlap. Thus, the overlap test is contorted into a search forx*, shown to be the solution of a nonlinear optimization problem that is easily solved once an associated scalar Lagrange multiplier is known. A successive approximations iteration equation for λ is obtained and is shown to converge as a contraction mapping. The method was developed to detect failures in an inertial navigation system that appear as uncompensated gyroscopic drift rate. For simulated gyroscopic failures, the iterations converged very quickly, easily allowing real-time failure detection.
Conference Paper
Full-text available
An alternative approach for the propagation of uncertain geometric information, based on the ideas presented by J.R. Deller (IEEE ASSP Magazine, vol.6, p.4-20, Oct. 1989) and extended to deal with graphs of geometric constraints, is presented. This method avoids the independency assumption of the probabilistic approach. In this approach, when new sensory data are acquired, a set of strips is obtained, propagated, and fused to obtain the updated ellipsoids associated with each feature, Then, the hypothesis about the location of the involved geometric features can be easily updated. Inconsistencies are easily detected, resulting in fast rejection of erroneous data
Conference Paper
A geometric approach is presented to solve data fusion problems. The approach uses bounded-error data parameter estimation rather than the usual statistical approach. Updating the location (orientation and position) of a mobile robot in a known polygonal environment is shown as example. The obtained recursive algorithms are similar to those of the Kalman filter with the advantage that only measures improving the estimates are considered
Conference Paper
We consider the problem of estimating the n-dimensional state of a dynamic system based on m-dimensional discrete-time measurements. The measurements depend nonlinearly on the state and are corrupted by white non-Gaussian noise. The problem is solved by recursively calculating the complete posterior density of the state given the measurements. For that purpose, a new exponential type density is introduced, the so called pseudo Gaussian density, which is used to represent the complicated non-Gaussian posterior densities resulting from the recursion. For polynomial measurement nonlinearities and pseudo Gaussian noise densities, it is shown that the result of the optimal Bayesian measurement update is exactly obtained by a Kalman filter operating in a higher dimensional space. The resulting filtering algorithms are easy to implement and always guarantee valid posterior densities
Article
In this paper we study the problem of estimating a given function of a vector of unknowns, called the problem element, by using measurements depending nonlinearly on the problem element and affected by unknown but bounded noise. Assuming that both the solution sought and the measurements depend polynomially on the unknown problem element, a method is given to compute the axis-aligned box of minimal volume containing the feasible solution set, i.e. the set of all unknowns consistent with the actual measurements and the given bound on the noise. The center of this box is a point estimate of the solution, enjoying useful optimality properties. The sides of the box represent the intervals of possible variation of the estimates. It is shown how important problems, like parameter estimation of exponential models, time series prediction with ARMA models and parameter estimation of discrete time state space models, can be formalized and solved by using the developed theory.
Article
For a large class of bounded-error estimation problems, the posterior feasible set for the parameters can be defined by nonlinear inequalities. The set inversion approach combines classical interval analysis with branch-and-bound algorithms to characterize . Unfortunately, as bisections have to be done in all directions of the parameter space, this approach is limited to problems involving a small number of parameters. Techniques based on interval constraint propagation make it possible to drastically reduce the number of bisections. In this paper, these techniques are combined with set inversion to bracket between inner and outer subpavings (union of nonoverlapping boxes). When only interested in the feasible intervals for the parameters, the set inversion approach becomes inefficient, and a new algorithm able to compute these intervals is given. This algorithm uses a new interval-based local research to compute the smallest box that contains . It is then compared with existing methods on an example taken from the literature.
Article
In the context of bounded-error estimation, one is interested in characterizing the set of all the values of the parameters to be estimated that are consistent with the data in the sense that the errors between the data and model outputs fall within prior bounds. While the problem can be considered as solved when the model output is linear in the parameters, the situation is far less advanced in the general nonlinear case. In this paper, the problem of nonlinear bounded-error estimation is viewed as one of set inversion. An original algorithm is proposed, based upon interval analysis, that makes it possible to characterize the feasible set for the parameters by enclosing it between internal and external unions of boxes. The convergence of the algorithm is proved and the algorithm is applied to two test cases. The results obtained are compared with those provided by signomial analysis.