Disturbance Decoupling of Nonlinear MISO Systems B
Disturbance Decoupling of Nonlinear MISO Systems B
Disturbance Decoupling of Nonlinear MISO Systems B
net/publication/229022321
CITATIONS READS
3 37
3 authors, including:
Claude H. Moog
Institut de Recherche en Communications et Cybernétique de Nantes
301 PUBLICATIONS 5,089 CITATIONS
SEE PROFILE
Some of the authors of this publication are also working on these related projects:
All content following this page was uploaded by Claude H. Moog on 22 May 2014.
Terms of use:
© Institute of Information Theory and Automation AS CR, 2002
Institute of Mathematics of the Academy of Sciences of the Czech Republic provides access to digitized
documents strictly for personal use. Each copy of any part of this document must contain these
Terms of use.
This paper has been digitized, optimized for electronic delivery and stamped with
digital signature within the project DML-CZ: The Czech Digital Mathematics Library
http://project.dml.cz
K Y B E R N E T I K A — V O L U M E 3 8 ( 2 0 0 2 ) , N U M B E R 5, P A G E S 6 0 1 - 6 0 8
This paper highlights the role of the rank of a differential one-form in the solution of
such nonlinear control problems via measurement feedback as disturbance decoupling prob
lem of multi-input single output (MISO) systems. For the later problem, some necessary
conditions and sufficient conditions are given.
1. INTRODUCTION
The disturbance decoupling problem by state feedback has been addressed in the
literature by several authors (see e. g., the historical accounts of [3, 6, 8]). In general,
most of the solutions assume that the state is available for measurement. When this
is not the case, two approaches may be followed: the reconstruction of the state by
means of an observer or the use of output feedback. Our objective is to investigate
solutions with the second kind of controller since it avoids approximations in the
estimated state and circumvents the superposition of controller and observer.
The few contributions which deal with the nonlinear disturbance decoupling prob
lem via output or measurement feedback are now briefly mentioned. The disturbance
decoupling problem has been solved by a geometric approach in [7]. The same prob
lem has also been considered in [9], and a necessary and sufficient condition has been
given using algebraic tools. For the multi-input multi-output (MIMO) case, there
just exists a sufficient condition in [1].
The purpose of this paper is to solve the disturbance decoupling problem via
measurement feedback for the class of MISO systems with a new approach based
on the rank of a differential one-form [4]. In the mean time through these results,
it allows to highlight the role of the rank of a differential one-form in the solution
of nonlinear control problem. For the disturbance decoupling problem via static
measurement feedback of MISO systems, some necessary conditions and sufficient
conditions are given. A complete solution is provided for a restricted class of such
systems.
The paper is organized as follows. In Section 2, some definitions and technical
tools useful are presented. Section 3 gives the main results of this paper and finally,
602 R. POTHIN, C.H. MOOG AND X. XIA
Consider,
The subspace fi is instrumental for solving the disturbance decoupling problem and
may be computed as the limit of the following algorithm:
Condition (i) represents the noninteraction with the disturbances whereas condition
(ii) represents the output controllability. Invertibility of (3(z) implies the existence
of an inverse matrix P~x(z) whose entries belong to K. Both /3(z) and f3~x(z) may
have singularities in Mn.
The following notations are borrowed from [4]. The originality of this paper
is mainly due to the fact that this definition is used for solving the disturbance
decoupling problem by static measurement feedback.
for some & G /C. In the sequel, we will call the set { d a i , . . . , da7_|_i} a basis of u.
As shown in [2, 4], the Pfaff-Darboux Theorem can be used to construct such a
basis. Another matrix based construction is described in [4].
The rank of a differential form will be extensively used to give necessary and suffi-
cient conditions for the disturbance decoupling problem with measurement feedback.
3. MAIN RESULTS
A direct extension of the result in [9] allows to give a sufficient condition for the
solvability of the problem.
(i) dy^ en + z + u.
(ii) There exists a one-form u G Z + U s.t.dy^ — u G ft and s.t. rank (u>) = 7 <
m — 1.
(iii) For any basis span^ {dai,. • • , d a 7 + i } of ft, i.e., a; = /3idai(z, u)+/32da2(z,u)+
P3da3(z, u) + ... + /? 7 + ida 7 +i (z + u),
z = Xз.
Example 2. This example show that for the MISO case condition (i) of Theo
rem 1 is not necessary. Indeed, this system can be disturbance decoupled by static
measurement feedback whereas condition (i) is not fulfilled.
Consider,
±1 = x2
x2 = x\ (sin 0:3)111 +x2(zosx3)u2 + x±
X3 = f(x)+q (6)
±4 = (C0SX3)U2
y = xi
z = X3.
Disturbance Decoupling of Nonlinear MISO Systems by Static Measurement Feedback 605
v>i = ^i/sinx3
^2 = i^/cosa^.
Theorem 1 becomes a necessary and sufficient condition for the class of SISO systems.
Moreover for this special class of system condition (iii) is always fulfilled. Then, the
following corollary is equivalent to the result in [9].
(i) dy ( r ) e n + Z + U.
Then (7) and (8) imply condition (i). Since u = ^d(F~1(z,u)), condition (ii) is also
fulfilled.
Since r is the relative degree of the output y, when condition (ii) is fulfilled, one can
set Xdv = u. Thus one obtain,
dy ( r ) Gflespanyjd*;}. (10)
On the other hand, one can see that for the class of MISO systems, which fulfills
the assumptions such that (7 fl Z = 0 and dy ( r ) G f i 0 2 + W, one has a necessary
and sufficient condition which can be written as follows:
606 R. POTHIN, C.H. MOOG AND X. XIA
(i) There exists a one-form UJ E Z + U s.t. dy^ - LJ E ft and s.t. rank (LJ) = 7 <
m — 1.
(ii) For any basis span A : {dai,..., d a 7 + i } of UJ, i. e., LJ = pidai + (32da2 + ... +
/?7+ida7+i
rank — [a{] = 7 + l(i = 1 , . . . , 7 + 1). (11)
Since one has dy(r) E fi, ® Z + U, this implies that UJ e Q, + Z. Assume that
UJ = UJ0 + UJZ for some (Do £ ft and UJZ e Z, then from the fact that
dy (r+*) G # + S pan{dv,...,dt; ( * ) },
That is, UJZ E ft, by the definition of ft. Thus cO2 = 0, due to the assumption that
$ln 2 = 0.
Then define UJ = £da(z,u), the necessity of (i) is fulfilled.
Prom [4] UJ can be described as
UJ = Pidai(z,u) + (32da2(z,u)
(12)
+ p3da3(z,u) + ... + /3 7 + ida 7 + i(z,w).
We show that for any such a choice, condition (ii) is necessarily fulfilled.
If not, there will be a linear combination
fidai + . . . + f 7 + i d a 7 + i E Z.
Assume without loss of generality that £1 7-= 0, then UJ can be decomposed into
in which
ůz = Pi
-^(Éidai + ... + f 7 + 1 d a 7 + i ) e Z
and
ßi
i = ßi- -r(ţi),
Çl
for i = 2 , . . . ,7 + 1.
Analogous to the proof of (i), we can show that uz = 0, a contradiction.
Thus (ii) holds.
E x a m p l e 3. Consider,
x\ = x2
x2 = xs (sin x2) + (cos x2)u
X3 = f(x)+q (13)
y = xi
Z = Xs.
Example 4. Consider,
±i = x2
±2 = x\ (sin xs)u\ +x2 (cos #3)^2
xs = f(x)+q (14)
y = xx
z = x3.
4. CONCLUSION
The notion of the rank of a one-form is used to solve the nonlinear disturbance
decoupling problem for MISO systems. At this moment, this new approach gives a
complete solution for a special class of MISO nonlinear systems. Further research
will consider a more general class of system. Moreover, note that this notion of rank
of one-form can be used for some other nonlinear control problem.
REFERENCES
[1] R. Andiarti and C. H. Moog: Output feedback disturbance decoupling in nonlinear
systems. IEEE Trans. Automat. Control 41 (1996), 1683-1689.
[2] R. L. Bryant, S. S. Chern, R. B. Gardner, H. L. Goldschmidt, and P. A. Griffiths: Ex-
terior Differential Systems. Springer-Verlag, New York 1991.
[3] G. Conte, C. H. Moog, and A. M. Perdon: Nonlinear Control Systems: An Algebraic
Setting. (Lecture Notes in Control and Information Science 242.) Springer-Verlag,
London 1999.
[4] H. Hermes: Involutive subdistributions and canonical forms for distributions and con-
trol systems. In: Theory and Applications of Nonlinear Control Systems (C. I. Byrnes
and A. Lindquist, eds.), North-Holland, Amsterdam 1986, pp. 123-135.
[5] H. J. C. Huijberts, L. Colpier, and P. Moreau: Nonlinear input-output decoupling by
static output feedback. In: Proc. 3rd European Control Conference (EEC'95), Rome
1995, pp. 1057-1062.
[6] A. Isidori: Nonlinear Control Systems. Second edition. Springer-Verlag, Berlin 1989.
[7] A. Isidori, A. J. Krener, C. Gori Giorgi, and S. Monaco: Nonlinear decoupling via
feedback: a differential geometric approach. IEEE Trans. Automat. Control 26 (1981),
331-345.
[8] H. Nijmeijer and A. J. van der Schaft: Nonlinear dynamical control systems. Springer-
Verlag, New York 1990.
[9] X. Xia and C. H. Moog: Disturbance decoupling by measurement feedback for SISO
nonlinear systems. IEEE Trans. Automat. Control 44 (1999), 1425-1429.
Dr. Richard Pothin, IRCCyN, UMR C.N.R.S. 6597, 1 rue de la Noe, BP 92101, 44321
Nantes Cedex 3. France.
e-mail: Richard.Pothin-RenexterQrenault. com
Dr. Claude Moog, IRCCyN, UMR C.N.R.S. 6597, 1 rue de la Noe, BP 92101, 44321
Nantes Cedex 3. France.
e-mail: [email protected]
Dr. Xiaohua Xia, University of Pretoria, Department of Electrical and Electronic Engi-
neering, Pretoria 0002. South Africa,
e-mail: [email protected]