Professional Documents
Culture Documents
Pomet1992 Explicit Design of Time-Varying Stabilizing Control Laws For A Class of Controllable Systems Without Drift
Pomet1992 Explicit Design of Time-Varying Stabilizing Control Laws For A Class of Controllable Systems Without Drift
North-Holland
* This work was done while the author was visiting at I N R I A ~, ukf~(x)
Sophla-Antipolls, 06565 Valbonne, France k=l
is &fferent from Aeo for any A in I~. Therefore, involving at most one f~ different from )'t, and
the map only once:
way to derive some stabilizing control laws with- where ~b is given by (5), /s well defined from
out solving any differential relations, except pos- ~( ~ n to [0, -~- 00) and enjoys the following prop-
sibly for finding some coordinates in which this erttes :
assumption is met: the control laws are actually (a) It ts 2rr-periodic with respect to ttme:
algebraic functions of the components of the vec-
V(t + 2w, x) = V(t, x). (11)
tor fields in these coordinates and of some vari-
ables and functions to be chosen by the designer. (b) V as well as its derivative with respect to x
vamsh if and only if x = O:
2.1. A general expression under Assumption 1
V(t, x ) = 0 ~ x = 0 , (12)
The following proposition gives a periodic OV
time-varying function, used in the design of the -~x(t,x)=0 ¢~ x = 0 . (13)
control laws, and meant to be a Lyapunov func-
tion for the final closed-loop system. (c) It is (the only) solution of the following
For a given time-varying function a(t, x), we hnear PDE with inttial conditions at t = 0:
may consider the time-varying ordinary differen- OV OV
tial equation - ~ ( t , x) + a ( t , x)-~x (t, x ) f l ( x ) =0, (14)
2 = a(t, x ) f , ( x ) , (4)
v(0, x) = ½ll xl[ 2, (15)
and define its 'general' solution ~b(t~, t2, x): it is
(d) For any positive real number K, the set
the value at time t I of the solution of (4) which
started from x at time t 2. In other words, { x, V( t, x) < K for some t ~ R} ts bounded.
(16)
OA(tl, t2, X)
Ot 1
Remark 2. (a) In the expression of V (10), the
= a ( t , , ~b(tl, t2, x))fl(~b(t ,, t 2, x)), (5a) squared norm might be replaced by any proper
~b(t, t, x) = x . (5b) positive function, vanishing only at 0, and with no
other singular point.
is defined, from classical results on ordinary (b) Majoration (9) might be replaced by any
differential equations, on an open subset of ~ × condition ensuring that all the solutions of .f =
x En containing the points (t 1, t 2, x) such that a(t, x)fl(x) are defined for any time.
t I = t 2. (c) Equation (7) might be replaced by
Proposition 1. Let a be a time-varying functton a(T-t, x) = - a ( t , x)
(i.e. a function from ~ × ~n to ~), and let us
for a certain T ~ R, or by any condition ensuring
suppose that:
that all the solutions of 2 = a(t, x)f~(x) are 2-rr-
(i) it is 2w-periodic and 'odd' with respect to
periodic.
time, and vanishes for x = O:
a(t+2"rr, x)=a(t,x) V(t,x), (6) This proposition will be proved in Section 4. It
a(-t,x)=-a(t,x) V(t, x), (7) turns out that, for any a meeting (6), (7), (8), (9),
the 2rr-periodic time-varying function 1/given by
a(t,0)=0 Vt; (8) (10) may be assigned to be non-increasing along
(ii) it satisfies the following majoration, for a the closed-loop system by using the following
certam K > 0: 2w-periodic time-varying control law:
Under Assumption 1 and a further reqmre- relation. It turns out that if the vector field ]i is
ment on a, this control law asymptotically stabi- globally redressable, it is actually possible to gwe
lizes (1). An explicit a meeting all the require- an explicit expression of V and the control laws;
ments may very easily be gwen, but we keep some it is the purpose of th~s section. As seen from
freedom on a in order to be able, on some example (36) in Section 3.2, it may often be
examples, to derive the simplest control laws we simpler, ff possible, to make the necessary trans-
can. This is made precise by the following theo- formations (i.e. a preliminary static feedback and
rem: a change of coordinates) to meet Assumption 2
than to the use the general expressions given in
Theorem 1. (a) Under Assumption 1, and if a is previous Section 2.1.
chosen such that conditwns (6), (7), (8), (9) of
Proposition 1 are satisfied and Assumption 2 (simplifies the expression of the
feedback law). There exists a system of coordi-
LfV(t,x)=O Vk~[1,m]]
nates (x~ . . . . . x , ) such that
= x=0, (18)
0Ja(t, x)=0 Vj>I
Otj f, - (21)
0x i
then 0 is a globally uniformly asymptottcally stable
equilibrium point of the time-varymg penodic Under this assumption, it is always possible to
closed-loop system obtained by applying the feed- choose V of the form
back law defined by (17), and the function V ts
non-increasing along its solutions. V ( t , x) = ~_(x
l 1 + h ( t , x 2 . . . . . x~)) 2
(b) It is always possible to find a function a
meeting the requtred condtttons. A possible chotce + ~1x 22 + • - • + ~x;,,
1 ?
(22)
lS
with some requirements on h, so that, instead of
JI x II 2 taking a as a design p a r a m e t e r and solving the
x)= sin t. P D E (14)-(15) for V, we are now able to take h
(1 + [[x[12)(1 + I l f l ( x ) I [ 2)
as a design p a r a m e t e r and compute a so as to
(19) satisfy (14)-(15):
ing theorem, which is a straightforward conse- Since [fl, f2] = a/ax2, we have
quence of Theorem 1:
Rank{f1, f2, [f~, f2]} = 3,
Theorem 2. (a) Under Assumption 1 and 2, and tf
so that Assumption 1 is satisfied. Assumption 2 is
h is chosen such that (23) and (24) are met and:
met as well in the original coordinates, so that we
=0 Vk ~ [2,m] may follow the method of Section 2.2 and take a
LfW(x) 1 ~ x=O. V of the form (22):
OJh
Vj> 1
0t'(t' x)=0
J (26)
V ( t , x ) = l ( x I + h ( t , x2,
1 2 1 2
x3)) 2
(32)
" - ~ X 2 q- ~X 3.
= r l f 1 4- t ' 2 f 2 (43)
so that Assumption 1 is met w~th f~ =g~ and x2 = -63 + cos so4 - 1, (45b)
f2 =g2 but not with f l =g2 and f2 =g~- Unfortu- X 3 4- i x 4 = -61 + i~:2
nately Assumption 2 is not met by f l = g~ since it
may be proven that the vector field gl is not - - e ~(~+c°s~4 l)f~'te-lC°~'r COS r dr.
globally redressable. It is therefore not possible /
~0
to use the method of Section 2.2 with fl =g~, (45c)
f2 =g2. The method of Section 2.1 does apply.
J - B Pomet / Ttme-varymg stabthzmg control laws 153
Note that finding these coordinates involved the Samson [10] designed some different control
expression of the solutions of laws for the same system, using
- x,- + x3 + cos t,
and
paper, and (52) is only a way of r e w r m n g the expression (58) because it turns out that taking
same controls. However, V given by (53) may be the norm cancels out the parts of 0 whmh cannot
checked to be the unique solution of be expressed simply
i~l'
- - + o~IL~,,V + ce2Lq,eV = 0, (57a) 3.3. A n extenston o f the present method
then the control law (62) yields global asymptotic To prove the other properties, let us explain
stability of the origin. where formula (10) comes from, and for which
This extension is interesting for two different reasons it makes V a solution of the PDE (14).
reasons: This is just an application of the method of
(a) In the context of this paper (systems with- characteristics for solving the linear PDE (14):
out drift meeting Assumption 1), it gives some (14) exactly means that the function V is constant
more degrees of freedom. As an aside conse- along the vector field O/at + af~ in ~ x Rn. This
quence, it allows one to interpret the control laws vector field is complete from (9) and its flow at
proposed by Samson [10] in the framework of this time r is the diffeomorphism of ~ × ~" given by
paper, see Section 3.2.
(t, x) ~ q,(t + ~, t, x ) .
(b) It may also extend to the more general
case of controllable systems without drift, not Formula (10) therefore states that V(t, x) is equal
necessarily meeting Assumption 1. This exten- to V(0, ~) where (0, ~) is the image of (t, x)
sion, using results from [3], is developed and under the flow of a / a t + a f ! at time - t , i.e.
discussed in the forthcoming [4]. Let us briefly (0, ~) is the only point on the same integral curve
outline it. It is not easy to deduce from controlla- of O/at + af~ as (t, x) with a zero t-component.
bility (full rank Control Lie Algebra) an explicit V is therefore obviously constant on the integral
expression of some a,'s meeting (63), whereas curves of O/Ot + o f 1, and this proves (14). Note
Assumption 1 allowed us to give the explicit that (14) may also be checked by hand: by differ-
expression (19) for a. The central difficulty solved entiating the identity
by Coron [3] is to establish existence of some a,'s
such that (his notations are not the same as here): t/,(~r, ~', ~(~', or, ~)) = ~: (64)
The proof of Proposition 2 and Theorem 2 are a ( z , $(~-, or, ~))fl(~b(z, iT, ~:)).
left to the reader. All the properties in Proposi-
Applying the resulting identity for or = 0, ~"= t
tion 2 can be checked by hand from (21), (22) and
and ~b(r, tr, ~) = x (i.e. ~: = ~O(0, t, x)), one gets
(25). Theorem 2 is actually a particular case of
Theorem 1 since it is easy to verify that a given 0q,
by (25) satisfies all the properties required in a~b(O, t, x) + (0, t x ) a ( t , x ) f l ( x ) =0.
'
established, one may notice that (81) is very simi- To prove that 0 is a unformly asymptotically
lar to the 'ad-condition' used in [6], and this stable equilibrium point of the time-variant
allows to make almost the same application of closed-loop system, or equivalently that the in-
LaSalle's invariance principle as in [6]. However, variant set S l × {0} (or the periodic orbit 0 = 1,
since it is somehow modified (one extra-dimen- x = 0) of system (69) is globally asymptotically
sion, lnvariant set S 1 × {0} instead of equilibrium attractive, we will use LaSalle's invariance princi-
point, a vanishes ...), we give the complete ple [8].
proof rather than referring the reader to [6]. Consider a solution of this closed loop system,
with initial condition (00, %). Since V decreases
Part (b) of the theorem is obvious: one may along this solution, it stays in the set
check by hand that a given by (19) meets all the V-I(V(Oo, Xo)), which is compact from (16). Any
requirements of part (a). Let us proceed with the solution is therefore bounded.
proof of (b). LaSalle's invariance principle states [8, Theo-
Since the functions a, V, u~ ..... u m from R × rem 2] that, since V decreases along any solution,
R" to R are 2w-periodic with respect to time, any bounded solution converges to the largest
they induce naturally some functions from S x × invanant set included in
~" to ~, the circle S 1 being the quotient A = {(0, x), v(0, x) = o}
S 1= ~/2w~. (68) :{X~S'X~", L F y ( X )=O, k = l . . . . . m}
(74)
The 2"rr-periodic time-variant closed-loop system
J.-B Pomet / Ttme-varymgstablhzmg control laws 157
(where the second equality comes from (73)), span the subspace {0} X TxRn of T(o,x)(Sl X ~"),
which is also the reunion of all the integral curves and from (80) they are in the kernel of the linear
of the vector field F which stay in A. form dV(0, x); this implies that dV is colinear to
From (17) and (74), if (0, x) e A then dO at this point, which is impossible from (13).
This actually proves that, on an integral curve
Ul(O, x) = a(O, x), (75a)
of the closed-loop system which stays in A, F 0
U2(0 , X) = 0 , . . . , Um(O, X) = 0, (75b) coincides with O/a0. Therefore, from (74) and the
fact that a is identically zero on this trajectory,
so that the vector field F coincides on A with the we have, at any point (0, x) of this trajectory,
vector field
~)Jot
3 at j (0, x ) = 0 , j>0, (82a)
Fo = -~ + aF 1. (76)
L/V(O, x) = 0 , k = 1 . . . . . m, (82b)
We therefore have to prove that the only integral
curve of F 0 (i.e. solution of X'= Fo(X)) which which is equivalent to x = 0 from (18). This proves
stays in A i s d = l , x = 0 . that any solution of the closed-loop system which
From (74), the functions LFV are identically stays in A statisfies x = 0, and completes the
zero on these trajectories, and so are their deriva- proof from LaSalle's invariance principle. []
tives: at any point X = (0, x) of such a trajectory,
Proof of Lemma 1. A simple inductive computa-
LJFoLFkV(X)=O, k = l . . . . . m;j>__O. (77)
tion from (76) shows that
Now, from (14), V is a first integral of F0:
ad~oFk = a~adlFF~
LFoV(X ) = 0 V X ~ - S 1X ~n, (78) + linear combination of
and it is not difficult to prove by induction on k ad~ Fk,, O<_j'<j-1;l<k'<m,
that this implies (83)
L ~FoLFy= LadkovkV. (79) and this proves the lemma. []
lng out to him the power of time-varying feed- brae-varying control law,, tor controllable systems wLthout
back laws for stabilization, demonstrated on some drift, Preprmt (19911
[5] L R Hunt, R Su and G Meyer, Design for multHnput
mobde robots. The author wishes also to thank nonhnear systems m R W Brockett, R S Mdlman and
Laurent Baratchart, from I N R I A Sophia-Antipo- H J Sussmann, Eds. Dtfferenttal Geometrw ( 'ontrol The~
hs, France, for some frmtful discussions on the or':' (Blrkhauser, Basel-Boston, 1983) 268-2ct8
subject and Jean-Michel Coron from Universlt6 [6] V Jurdjevlc and J P Qumn, Controllabdlty and stabdlty,
Paris-Sud, Orsay, for suggesting condition (7) to J Dtf]erenttal Equatton~ 28 (1978) 381-389
[7] B Jakubczyk and B Respondek, On Ilneanzatlon of
ensure periodicity of all the solutions of (4). control systems, Bull Acad Polonat~e Sol Ser Su Math
28 (19801 517-522
[8] J-P LaSalle, Stabdlty theory for ordinary differential
References equahons, J Dtfferenttal Equattons 4 (1968) 57-65
[9] C Samson, Velocity and torque feedback control a non-
holonomlc cart, Internat Workshop m Adapttt'e and Non-
[1] R.B Brockett, Asymptotic stabdtty and feedback stabl-
hzatlon, m R W Brockett, R S. Mdlman and H J Suss- hnear Control Issue~ m Robottcs, Grenoble (Sprmger-
Verlag, Berhn-New York, 199(I)
mann, Eds., Dtfferenttal Geometrtc Control Theory (Birk-
hauser, Basel-Boston, 19831 181-191. [10] C. Samson, Time-varying stabdlzatlon of a nonholonomlc
car-hke mobde robot, Report INRIA Sophla-Antlpohs,
[2] G Campion, B. d'Andr~a-Novel and G Bastm, Control-
lablhty and State-Feedback Stabdlzatlon of Non Holo- (19901
[11] R S6pulchre, Preprmt, Umv Cathohque de Louvam,
nom~c Mechamcal Systems, lnternat. Workshop tn Adap-
twe and Nonhnear Control Issues m Robottcs, Grenoble Belgmm (1991)
[12] E D Sontag, Feedback stabdlzatlon of nonhnear systems,
(Sprmger-Verlag, Berhn-New York, 1990)
[3] J -M Coron, Global asymptohc stabdizatlon for control- tn M A Kaashoek et al, Eds., Robust Control of Lmear
Systems and Nonhnear Control Proceedmgs of the Interna-
lable systems without drift, Preprmt (1991), to appear m
tional Symposium MTNS-8q (Blrkhauser, Basel-Boston,
Math Control Stgn Systems
[4] J -M Coron and J.-B Pomet, A remark on the design of 19901 61-81