HAL Id: hal-00982014
https://hal.archives-ouvertes.fr/hal-00982014
Preprint submitted on 23 Apr 2014
HAL
is a multi-disciplinary open access archive for the deposit and dissemination of sci- entific research documents, whether they are pub- lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers.
L’archive ouverte pluridisciplinaire
HAL, estdestinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.
Nonlinear observer based on observable cascade form
Mariem Sahnoun, Hassan Hammouri
To cite this version:
Mariem Sahnoun, Hassan Hammouri. Nonlinear observer based on observable cascade form. 2014.
�hal-00982014�
Nonlinear observer based on observable cascade form
Mariem Sahnoun and Hassan Hammouri
Abstract— In this paper, the error observer linearization is extended to a class of observable cascade systems which contains state affine systems up to output injection. First, we give a theoretical result which states necessary and sufficient conditions. Next, we give an algorithm permitting to calculate a system of coordinates in which a nonlinear system takes the desired cascade observable form.
Index Terms— Nonlinear systems, output injection, nonlinear observer.
I. INTRODUCTION
The implementation of linear or nonlinear observers in control systems design, fault detection and other domains is well understood by now.
To design an observer for nonlinear systems, many ap- proaches have been developed. Among them, the geometric approaches consist in characterizing nonlinear systems which can be transformed by a change of coordinates to a special class of systems for which a simple observer can be designed.
The observer error linearization problem consists of trans- form a nonlinear system into a linear one plus a nonlinear term depending only on the known inputs and outputs. For such systems, a Luenberger observer can be designed. This problem has attracted a good deal of attention, since its formulation by [9] (see for instance [2], [3], [10]–[13]. Using immersion technics, an extension of this problem has been stated in [8] in the single output case. In the same spirit as for the error linearization problem, the authors in [4]–[7]
characterized nonlinear systems which can be steered by a change of coordinates to state affine systems up to output injection. For these systems, a Kalman-like observer can be designed.
In this paper, we will characterize nonlinear systems which can be transformed by local coordinate systems into the following cascade form:
˙z=A(u)z+ψ(u,y) e˙
z=A(u)e ez+ψe(u,z,ey) Y=
y e y
= Cz
e Cez
(1)
For these systems, an observer structure may take the
M. Sahnoun and H. Hammouri are with Universit´e de Lyon, F-69622, Lyon, France; Universit´e Lyon 1, Villeurbanne;
CNRS, UMR 5007, LAGEP (Laboratoire d’Automatique et de G´enie des Proc´ed´es). 43 bd du 11 novembre, 69100 Villeurbanne, France sahnoun@lagep.univ-lyon1.fr, hammouri@lagep.univ-lyon1.fr
following form:
b˙
z=A(u)bz+ψ(u,y)−S−1CTR(Cbz−y) be˙z=A(u)e bez+ψe(u,bz,ey)−Se−1CeTR(eCebez−y)e
S˙=−θS−AT(u)S−SA(u) +CTRC
˙e
S=−θeSe−AeT(u)eS−SeA(u) +e CeTReCe
(2)
where S(0), S(0), R ande R are symmetric positive definitee matrices, θ >0, θe>0 are parameters. The proof of the convergence of this observer has been stated in [1].
This paper is organized as follows:
In section II, the problem under consideration is formalized and an existence theorem is stated. In section III, an al- gorithm permitting to calculate a system of coordinates in which a nonlinear system takes the desired cascade form is proposed.
II. PRELIMINARY RESULTS AND EXISTENCE THEOREM
A. Preliminary results
For the sake of simplicity, we only consider the case where the outputs y and ey are scalars. The following classes of nonlinear systems will be considered:
˙
x=f(u,x) y=h(x) e y=eh(x)
(3)
where x∈Rn, the input u(t)∈Rm and the outputs y(t)and e
y(t)are belong toR. f , h andeh are assumed to be of class C∞.
We adopt the following definition.
Definition 1: System (1) is said to be cascade- observable, if system (1) together with its associated reduced system in z are observable.
The following geometric notions will be used in the sequel.
In the system of coordinates(x1, . . . ,xn), let X=∑ni=1αi ∂
∂xi
be a vector field and let ω =∑ni=1aidxi a one-differential form, then the following operations will be considered:
• Lie derivative action: LX(ω) = ∑ni=1αiLX(ai)dxi+
∑ni=1aidαi
• The duality product:ω(X) =∑ni=1αiai
The above duality product can be extended to k- differential forms as follow:
If ω = ∑1≤i1<...<ik≤na(i
1,...,ik)dxi1 ∧. . .∧dxik is a k-differential form and X = (X1, . . . ,Xk) is a k- tuple of vector fields, with Xi =∑nl=1αil ∂
∂xl, then
ω(X) =∑1≤i1<...<ik≤na(i1,...,ik)
α1i1 . . . αki1
. . . . α1ik . . . αkik
.
• Inner product: Let X = (X1, . . . ,Xl) be a l-tuple of vector fields, with l≤k. Then iX(ω) is the (k−l)- differential form defined by:
iX(ω)(Y1, . . . ,Yk−l) =ω(X1, . . . ,Xl,Y1, . . . ,Yk−l).
In particular, if k=l, then iX(ω)is a function (a 0−differential form).
Let fube the vector field defined by fu(x) =f(u,x)and let X be a vector field onRn. We define the family of real vector spacesΩXk of 2-differential forms as follows:
• ΩX0 =0 andΩX1 =S pan{dLfu(h)∧dh; u∈Rm}. Notic- ing that these two spaces do not depend on X ,
• for k≥1, we set ΩXk+1=S pan{Lfu(iX(ω))∧dh; u∈ Rm; ω∈ΩXk}+ΩXk.
Now settingπ=dϕ1∧. . .∧dϕq, whereϕkareC∞functions, and letXe= (Xe1, . . . ,Xeq+1)be a(q+1)-tuple of vector fields.
As above, we define the vector spaces ΩXk,πe of (q+2)- differential forms as follows:
• ΩX0,πe =0 andΩX1,πe =S pan{dLfu(eh)∧deh∧π; u∈Rm},
• for k≥1, ΩXk+1,πe =S pan{Lfu(iXe(ωe))∧deh∧π; u∈ Rm; ωe∈ΩXk,πe }+ΩXk,πe .
B. Existence theorem
In the single output case (see [4], [6]), ( [5] for the the multi-output case) the authors gave necessary and sufficient conditions under which nonlinear systems can be trans- formed in a state affine system up to output injection.
The following theorem states an existence theorem which extends those stated in [4], [5]:
Theorem 1:
Observable system (3) can be transformed by a local change of coordinates around some x0∈Rn to a cascade- observable system (1) in which C andC are of rank 1, if ande only if, the following conditions hold on some neighborhood of x0:
1) It exists a vector field X satisfying the following con- ditions:
1-i) LX(h) =1.
1-ii) The algebraic sumΩX=
∑
k≥1
ΩXk is a real vector space of dimension q−1.
1-iii) For everyω∈ΩX, d(iX(ω)) =0.
1-iv) The dimension of [Vq−1(iX(ΩX))∧dh]|x0 is equal to 1, where[Vq−1(iX(ΩX))∧dh]|x0 ={iX(ω1)∧. . .∧ iX(ωq−1)∧dh(x0); ωi∈ΩX ,1≤i≤q−1}.
2) Consider the following functions ϕ1, . . . ,ϕq+1 defined by:
ϕ1=h ϕq+1=eh
(dϕ1, . . . ,dϕq)forms a basis of iX(ΩX) +Rdh (4)
Settingπ=dϕ1∧. . .∧dϕq, then there exists a(q+1)- tuple of vector fields Xe= (Xe1, . . . ,Xeq+1) satisfying the following conditions on some neighborhood of x0: 2-i) LXe
i(ϕj) =δi j, whereδi j=1 if i=j and 0 otherwise.
2-ii) The algebraic sum ΩXπe=∑k≥1ΩXk,πe is a real vector space of dimension n−q−1.
2-iii) For everyωe∈ΩXπe, d(iXe(ωe)) =0.
2-iv) The dimension of [Vn−q−1(iXe(ΩXπe))∧dϕ1∧. . .∧ dϕq+1]|x0 is equal to 1.
The proof of theorem 1 can be obtained by following the same approach as the one proposed in the works [4], [5].
The outline of the proof is summarized as follows:
1) Sufficient condition: iX(ΩX) and iXe(ΩXπe) are vector spaces of dimension q−1 and n−q−1 respectively, and (iX(ω1), . . . ,iX(ωq−1)), (iXe(ωe1), . . . ,iXe(ωen−q−1)) are their respective bases. Setting dz1 =dh, dzi = iX(ωi+1), dez1=deh and dezi=iXe(ωei+1). It can be shown that Lfu(zi) =
∑
q j=2ai j(u)zj+ψi(u,z1) and Lfu(ezi) =
n−q
∑
j=2
e
ai j(u)ezj+ψi(u,z,ez1). Consequently, in the(z,ez)sys- tem of coordinates system (3) takes the cascade form (1).
2) Necessary condition: Since conditions 1), 2) of theorem 1 are intrinsic (they do not depend on the system of coordinates), it suffices to show them for the cascade observable system (1). After a simple linear change of coordinates, we can assume that y=Cz=z1 and ey= Ceez=ez1, and it can be shown that X= ∂
∂z1
and Xe= (eX1, . . . ,Xeq+1) = ( ∂
∂z1
, . . . , ∂
∂zq
, ∂
∂ez1)satisfy conditions 1) and 2) of theorem 1.
In the following, we focus on the development of an algorithm permitting to calculate vector fields X,Xe1, . . . ,Xeq+1 which meet conditions 1)and 2)of theorem 1.
III. PROCEDURE OF CALCULATION OF VECTOR FIELDS
X,Xe1, . . . ,Xeq+1 A. Preliminary results
The following notations will be used in the sequel:
• Let V be a vector space, and W a subspace of V , then forξ,ξ′∈V , the notation ξ =ξ′ modulo (W) means thatξ =ξ′+w, for some w∈W .
• Setting F (resp. V) to be a set of one-differential form (resp. of vector fields). D=S pan(F) (resp.∆= S pan(V)) will denote the co-distribution (resp. the distribution) spanned byF (resp. byV).
• The orthogonal of a co-distribution D is the distribution
∆=Ker(D) =S pan({X ; ω(X) =0, ∀ω∈F}), where ω(X) is the duality product between one-form and vector fields. In particular, ifF is spanned by a family of one-exact form {dϕ; ϕ ∈Ff}, then ∆=Ker(D)
is the distribution spanned by the set of vector fields {X ; LX(ϕ) =0, ∀ϕ∈Ff}.
• Let D, D′be two co-distributions, with D′⊂D, then the quotient D/D′will denote the set of equivalent class of differential forms[ω] =ω+D′={ω+ω′; ω′∈D′}, whereω∈D. Similarly, if∆⊂∆′are two distributions, elements of the quotient∆′/∆will be denoted by[X] = X+∆ where X∈∆′.
If [ω]∈D/D′ andχ∈D such that[ω] = [χ], then we setω=χ modulo (D′).
Finally, if X , Z are two vector fields,[X,Z]will denote the Lie bracket of these vector fields.
The following flag of co-distributions and distributions will be considered:
D0⊂. . .⊂Dk⊂. . .
∆0⊃. . .⊃∆k⊃. . . De0⊂. . .⊂Dek⊂. . . e∆0⊃. . .⊃∆ek⊃. . .
(5)
Where,
• D0=0 the null co-distribution, D1=S pan({dh}), by induction Dk+1=Dk+S pan({dLfuk. . .Lfu1(h); u1, . . . , uk∈Rm}), and D♯=∑k≥1Dk.
• De0=D♯, De1=De0+S pan({deh}), for k≥1, Dek+1= e
Dk+S pan({dLfuk. . .Lfu1(eh); u1, . . . ,uk ∈ Rm}), and e
D♯=∑k≥1Dek.
• ∆k=Ker(Dk), and∆♯=Ker(D♯).
• e∆k=Ker(Dek), and∆e♯=Ker(eD♯).
• The quotient co-distribution Dk/Dk−1 (resp.Dek/Dek−1) is the dual of the quotient distribution ∆k−1/∆k (resp.
e∆k−1/e∆k). The duality product[ω]([X]) =ω(X)is well defined.
In the two following claims, fu=
∑
q i=1(Ai(u)z+ψi(u,y)) ∂
∂zi
+
n−q
∑
i=1
(Aei(u)ez+ψei(u,z,ey)) ∂
∂ezi, and the outputs h,eh are respec- tively y=Cz=z1,ey=Ceez=ez1.
Considering the ringsHk,Hfk such that:
• H0=C∞{z1} (resp. Hf0=C∞{z1, . . . ,zq,ez1}) is the ring ofC∞-functionsϕ(z1)(resp.ϕ(z1, . . . ,zq,ez1)).
• C∞{z} (resp. C∞{z,ez}) denotes the ring of C∞- functionsϕ(z1, . . . ,zq)(resp.ϕ(z1, . . . ,zq,ez1, . . . ,ezn−q)).
Then for k≥1, Hk (resp. Hfk) is the smallest sub- ring of C∞{z} (resp. of C∞{z,ez}) containingHk−1∪ {CA(u1). . .A(uk)z; u1, . . . ,uk ∈ Rm} (resp. Hfk−1∪ {CeeA(u1). . .A(ue k)ez; u1, . . . ,uk∈Rm}).
Then we have:
Claim 1:
i) Lfuk. . .Lfu1(Cz) =CA(u1). . .A(uk)z modulo(Hk−1).
ii) Lfuk. . .Lfu1(Ceez)) =CeA(ue 1). . .A(ue k)ez modulo(Hfk−1).
The following claim can be deduced from the above one.
Claim 2:
• The flags of co-distributions D0 ⊂. . . ⊂Dk ⊂. . .;
e
D0/D♯⊂. . .⊂Dek/D♯⊂. . . are of constant dimensions and
defined as follows:
a) D1=S pan(dCz), and for k≥2, Dk is spanned by the set of one-forms{dCz} ∪ {dCA(u1). . .A(ul)z; 1≤l≤ k−1, uj∈Rm}.
b) Similarly, De1/D♯ can be identified with the co- distribution S pan(dCeez), and for k≥2,Dek/D♯is isomor- phic to the co-distribution spanned by the set of one- forms {dCeez} ∪ {dCeA(ue 1). . .A(ue l)ez; 1≤l≤k−1, uj∈ Rm}.
• System (1) is cascade observable iff: dim D♯=q (q is the dimension of the z-space), and dimDe♯/D♯=n−q (n−q is the dimension ofez-space).
In the sequel, we set ν (resp νe) to be the smallest integer such that Dν=D♯ (resp.Deνe/D♯=De♯/D♯):
D0⊂. . .⊂Dν=Dν+1
De0/D♯⊂. . .⊂Deeν/D♯=Deeν+1/D♯
(6) This subsection will be ended by the two following technical results:
Lemma 1:
If dϕ∈Dk−1(resp. dϕe∈Dek−1) and X∈∆k−1(resp.Xe∈ e∆k−1), then dϕ([fu,X]) =−d(Lfu(ϕ))(X) = −LX(Lfu(ϕ)) (resp. dϕ([fu,X]) =e −d(Lfu(eϕ))(Xe) =−LXe(Lfu(ϕe))).
Proof of lemma 1.
Let dϕ ∈Dk−1 and X ∈∆k−1, the equality dϕ([fu,X]) =
−d(Lfu(ϕ))(X)follows from the following facts:
• dϕ([fu,X]) =Lfu(LX(ϕ))−LX(Lfu(ϕ))
=d(LX(ϕ))(fu)−d(Lfu(ϕ))(X),
• X∈∆k⊂∆k−1=Ker(Dk−1),
• LX(ϕ) =dϕ(X) =0
Similar argument can be used to prove dϕ([fu,X]) =e
−d(Lfu(eϕ))(Xe).
Claim 3:
Let Z = (Z1, . . . ,Zk) be a k-tuple of vector fields, let
g, ϕ1, . . . ,ϕk be C∞-functions such that dϕ1∧. . .∧dϕk is
nowhere vanish and that LZj(ϕi) =δi j, then:
iZ(dg∧dϕ1∧. . .∧dϕk) =dg−
∑
k j=1LZj(g)dϕj. More precisely, we have:
iZ(dg∧dϕ1∧. . .∧dϕk) = (−1)q(dg−
∑
k j=1LZj(g)dϕj).
B. Algorithm
In this subsection, we will give an algorithm permitting to calculate the vector fields X,Xe1, . . . ,Xeq+1, which meet conditions of theorem 1. This algorithm will be obtained in three steps:
1) The first step consists to calculate X using only f(u,x) and h(x).
2) The knowledge of f(u,x), h, eh(x) and X allows to calculateXeq+1.
3) Finally,Xe1, . . . ,Xeqcan be computed based on the knowl- edge of f(u,x), h,eh(x), X andXeq+1.
Assuming that the flags of co-distributions:
0=D0⊂. . .⊂Dν=Dν+1
0=De0/Dν⊂. . .⊂Deeν/Dν=Deeν+1/Dν (7)
are of constant dimensions and that dim(Dν) = q, dim(Deeν/Dν) =n−q.
For k≥1, we define the bases Bk and Bek of Dk/Dk−1 and e
Dk/Dek−1as follows:
B1={[dh]}, Be1={[deh]}
for k≥2 :
Bk={[d(Lfuk−1. . .Lfu1(h))]; (u1, . . . ,uk−1)∈Uk−1} e
Bk={[d(Lfuk−1e . . .Lf
eu1(eh))]; (ue1, . . . ,euk−1)∈Ufk−1} (8) for some subsetsUk−1 andUfk−1 of(Rm)k−1.
The symbol[(.)]stands for the equivalent class of(.).
Now, let B∗ν,Be∗eνbe the respective dual bases of Bν andBeeν (B∗ν,Be∗eν are bases of ∆ν−1/∆ν and∆eeν−1/e∆eν), the following vector fields will be required in theorem 2 below :
• The vector fields[Zu1...uν−1],[eZue1...eu
eν−1]:
Let (u1, . . . ,uν−1), (resp. (ue1, . . . ,ueν−1e )) be fixed elements ofUν−1 (resp. ofUf
eν−1), then[Y] = [Zu1...uν−1](resp.[eY] = [eZue1...ue
eν−1]) is the element of B∗ν (resp. of Be∗eν) defined by:
for(v1, . . . ,vν−1)∈Uν−1, d(Lfvν−1. . .Lfv1(h))(Y) =1, if (u1, . . . ,uν−1) = (v1, . . . ,vν−1), and 0 otherwise for(ev1, . . . ,eveν−1)∈Ufν−1e , d(Lfev
eν−1. . .Lfe
v1(eh))(Ye) =1, if (eu1, . . . ,ueeν−1) = (ve1, . . . ,veeν−1), and 0 otherwise
(9)
• The vector fields[Yu1...uν−1],[eYue1...ueeν−1]:
Setting [Y] = [Zu1...uν−1]and[eY] = [eZue1...ue
eν−1], then:
Yu1...uν−1 = [fuν−1,[. . . ,[fu1,Y]. . .]]]
e
Yue1...ueeν−1 = [fue
eν−1,[. . . ,[fue1,Y]. . .]]] (10) In order to state lemma 2 below, the following notations will be required:
• Let(dϕ1, . . . ,dϕq)be a basis of Dν and dϕq+1=deh.
• Settingπe=dϕ1∧. . .∧dϕq+1.
• LetXe= (Xe1, . . . ,Xeq+1)be a(q+1)-tuple of vector fields satisfying LXe
i(ϕj) =δi j.
• Foreu1∈Uf1, we set ωeeu1=dLfe
u1(eh)∧πe.
• For k ≥2 and (eu1, . . . ,uek)∈Ufk, we set ωeue1...euk = Lfe
uk(iXe(ωeeu1...euk−1))∧πe. Thus we have:
Lemma 2:
For 1≤k≤eν−1; for every(ue1, . . . ,euk)∈Ufkthe following
properties hold:
ωeue1...uek=dLf
uke . . .Lfu1e (eh)∧πe +
k−1
∑
l=1
∑
(ue1,...,uel)∈Ufl
gue1...ue
l(x)dLfeul. . .Lf
eu1(eh)∧πe (11) iXe(ωeue1...euk) =dLfeuk. . .Lfeu1(eh)−
∑
q j=1LXe
jLfeuk. . .Lfu1e(eh)dϕj+Θk
Θk=Θek−
∑
q j=1k−1
∑
l=1
∑
(eu1,...,eul)∈Ufl
geu1...uel(x)LXe
jLfe
ul. . .Lfe
u1(eh)dϕj
Θek=
k−1
∑
l=1
∑
(ue1,...,uel)∈Ufl
geu1...eul(x)dLfeul. . .Lfu1e(eh) +gk(x)dϕq+1
(12) with the property that gue1...uel(.), gk(.) are C∞-functions which do not depend on(eX1, . . . ,Xeq).
Proof of lemma 2.
• For k=1:
Let u1∈U1, by definition ωeue1 =dLfe
u1(eh)∧πe, and from claim 3, we know that iXe(ωeeu1) = dLf
eu1(eh)−
q+1
∑
j=1
LXe
jLfe
u1(eh)dϕj=dLfe
u1(eh)−
∑
q j=1LXe
jLfe
u1(eh)dϕj+Θ1, hereΘ1=LXe
q+1Lfe
u1(eh)dϕq+1. Hence (11), (12) are true for k=1.
• Assuming that (11), (12) hold for 1≤l≤k−1, and let us show them for k. Using the definition ofωeeu1...uek and applying (12) for k−1, we get:
ωeue1...uek=dLf
uke . . .Lfeu1(eh)∧πe
−Lfuke[
∑
q j=1LXe
jLfe
uk−1. . .Lfu1e (eh)dϕj]∧πe+Lfeuk(Θk−1)∧πe Θk−1=Θek−1−
∑
q j=1k−2
∑
l=1
∑
(eu1,...,eul)∈Uf
l
gue1...eul(x)LXe
jLf
eul. . . Lf
u1e (eh)dϕj
Θek−1=
k−2
∑
l=1
∑
(ue1,...,uel)∈Ufl
gue1...eul(x)dLfeul. . .Lfeu1(eh)
+gek−1(x)dϕq+1
(13) and gue1...eul, gek−1 do not depend on(eX1, . . . ,Xeq).
Using the fact that dϕi∈Dν, for 1≤i≤q, and that Lfu(Dν)⊂Dν, then the following equality holds for every smooth functions a1(x), . . . ,aq(x):
Lfu(
∑
q j=1aj(x)dϕj)∧πe=0 (14) Combining (14) with expressions of Θk−1, Θek−1, we get:
ωeue1...euk=dLfe
uk. . .Lfe
u1(eh)∧πe+
k−2
∑
l=1
∑
(ue1,...,uel)∈Uf
l
Lfe
uk
[egue1...eul(x)dLfeul. . .Lf
eu1(eh) +gek−1(x)dϕq+1]∧πe (15)
By construction Lfe
uk[geue1...uel(x)dLfeul. . .Lfe
u1(eh)] and Lf
uke(egk−1(x)dϕq+1)∧πe do not depend on (Xe1, . . . ,Xeq) and{dϕ1, . . . ,dϕq+1} ∪{dLfeul. . .Lfu1e (eh); (ue1, . . . ,eul)∈
f
Ul, 1≤l≤k−1} forms a basis of Dek, hence the last term of the right hand expression (15) takes the form
k−1
∑
l=1
∑
(ue1,...,uel)∈Ufl
gue1...uel(x)dLfeul. . .Lfe
u1(eh)∧πe, where the gue1...ue
l(x)’s are C∞-functions which do not depend on(Xe1, . . . ,Xeq). Consequently, expression (11) is satis- fied.
In order to end the proof of lemma 2, it remains only to check (12).
Applying claim 3 to expression (11), we get:
iXe(ωeue1...uek) =dLf
euk. . .Lf
eu1(eh) +
k−1
∑
l=1
∑
(eu1,...,eul)∈Uf
l
geu1...eul(x)
dLf
eul. . .Lf
eu1(eh)−
q+1
∑
j=1
LXe
jLf
uke . . .Lfeu1(eh)dϕj
−
q+1
∑
j=1 k−1
∑
l=1
∑
(ue1,...,uel)∈Ufl
gue1...ue
l(x)LXe
jLf
eul. . .Lf
eu1(eh)dϕj
(16) Finally, expression (12) follows from (16) in which we introduce:
Θek=
k−1
∑
l=1
∑
(eu1,...,eul)∈Ufl
gue1...eul(x)dLfe
ul. . .Lfe
u1(eh) +gk(x)dϕq+1
where gk(x) =−LXe
q+1Lfuke . . .Lfeu1(eh)−
k−1
∑
l=1
∑
(ue1,...,uel)∈Ufl
gue1...eul(x)LXe
q+1Lfule. . .Lfeu1(eh) Θk=Θek−
∑
q j=1k−1
∑
l=1
∑
(ue1,...,eul)∈Ufl
gue1...eul(x)LXe
jLfeul. . .Lfeu1(eh)dϕj
(17) Moreover, by construction geu1...uel(x) and gk do not depend on (Xe1, . . . ,Xeq). This ends the proof of lemma 2.
Now we can state the algorithm which allows to calculate vector fields X ,Xe1, . . . ,Xeq,Xeq+1satisfying conditions 1) and 2) of theorem 1.
Theorem 2: (Algorithm)
System (3) can be steered by a local change of coordinates around some x0 to a cascade-observable system (1), if, and only if, the following conditions hold:
a) The flag of co-distributions D0⊂. . .⊂Dν =Dν+1, De0/Dν⊂. . .⊂Deeν/Dν=Deeν+1/Dν are of constant di- mension on some neighborhood of x0, and dim(Dν) =q, dim(Deeν/D♯) =n−q
b) Let Bν and Beeν be any fixed bases of Dν/Dν−1 and Deeν/Deν−1e (see the construction (8)). Let Y and Y bee any fixed vector fields of the form[Y] = [Zu0
1...u0ν−1]∈B∗ν and[eY] = [eZeu0
1...eu0eν−1]∈Be∗eν, then the following properties hold:
1) The vector X= (−1)ν−1Yu01...u0ν−1 satisfies condition 1) of theorem 1.
2) Setting Xeq+1 = (−1)eν−1Yeeu01...eu0eν−1 and considering C∞-functionsϕ1, . . . ,ϕq+1such thatϕ1=h,ϕq+1=eh and that (dϕ1, . . . ,dϕq) forms a basis of iX(ΩX) + Rdh. Let Xe1, . . . ,Xeq be vector fields satisfying LXe
j(ϕi) =δi j, 1≤j≤q, 1≤i≤q+1, and such that for every(ue1, . . . ,euk)∈Ufk, 1≤k≤eν−1, we have:
∑
q j=1d(LXe
jLf
uke . . .Lfu1e (eh))∧dϕj=dΘue1...uek (18) where Θue1...euk is the one-differential form stated in (12). ThenXe1, . . . ,Xeq+1 satisfy condition 2) of theo- rem 1.
Remark 1: According to expression (12) of lemma 2, expression (18) is then equivalent to d(iXe(ωeue1...uek)) =0.
Some comments on the procedure of calculation of vector fields X , Xe1, . . . ,Xeq,Xeq+1:
1) The calculation of the vector field X requires only the knowledge of expressions of fuand h.
2) Xeq+1can be directly computed from the knowledge of X , fu, h andeh.
3) For 1≤i≤q+1, the functionsϕi can be deduced from X , fuand h andeh.
4) Finally, we end these comments by giving the algorithm of computation of(Xe1, . . . ,Xeq):
Computation of(Xe1, . . . ,Xeq):
Based on the construction of Bek and the functions ϕ1, . . . ,ϕq+1, the set {ϕ1, . . . ,ϕq+1} ∪ {Lfeuk. . .Lfeu1(eh); 1≤
k≤ eν−1, (ue1, . . . ,uek) ∈ Ufk} forms a local system of
coordinates, which we denote by(ξ,ξe), and where ξ= (ϕ1, . . . ,ϕq+1) = (ξ1, . . . ,ξq+1) ξe= (ξe1, . . . ,ξeeν−1), ξek= (ξek1, . . . ,ξek,de
k) where {dξek1, . . . ,dξek,de
k}={dLfeuk. . .Lf
eu1(eh); (ue1, . . . ,euk)∈ f
Uk}, and{[dLfeuk. . .Lfu1e(eh)]; (ue1, . . . ,uek)∈Ufk}=Bek+1. Therefore, we adopt the following notations:
ωeeu1...euk=ωeki=dξeki
LXe
j(ξeki) =LXe
jLf
euk. . .Lf
eu1(eh) =Xekij Using the fact that LXe
j(ϕj) = LXe
j(ξj) =δi j, we obtain e
Xj= ∂
∂ξj
+
eν−1 k=1
∑
dek
i=1
∑
e Xkij ∂
∂ξeki
. Thus, the expression (12) can be rewritten:
ie
X(ωeki) =dξeki−
∑
q j=1Xekijdξj+Θki (19) where theΘki’s are one-differential forms depending at most onXelij, 1≤l≤k−1, 1≤j≤q+1.
The calculation ofXekij’s follows from the following recursive procedure: