Domain of Attraction Remarks on the domain of attraction x f ( x ), f (0 ) 0, x R n R A { x 0 R : x ( x 0 , t 0 , t ) 0 as t } n Complete (total) domain of attraction Estimate of Domain of attraction : Rˆ A Lemma : The complete domain of attraction is an open, invariant set. Its boundary is formed by trajectories 5-1 Consider x f ( x ) Let D R n be such that Is D in Rˆ A ? V ( x) 0 and V 0, x D , x 0 V c1 V c2 c 2 c1 D V might be positive could escape from D What is a good Rˆ A ? Consider 2 2 2 V 0 for all x1 x 2 a V (x) C D * Rˆ A D c 5-2 Example Ex: x1 2 x1 x1 x 2 0 0 x 2 x 2 x1 x 2 1 2 0 Here is asymptotic ally stable since 0 Thus let Q I , PA A P I f x 14 P 0 T A x0 2 0 0 1 0 1 2 Since V ( x ) x Px T 2 2 2 2 V ( x ) ( x1 x 2 ) ( 12 x1 x 2 x1 x 2 ) x Here 2 1 2 x1 x 2 1 2 x 2, 2 x 2 2 x1 x 2 x1 2 x 2 2 5 4 x x1 2 x 2 3 2 2 x 2 (1 5 x 5 4 2 x 2) 5-3 Example (Continued) Thus V is negative Note that V c with Therefore definite in a ball of radius 2 x Px min ( P ) x 2 . Thus T r 4 5 . we can choose level set c min ( P ) r . 2 c 0 . 79 The set c with 1 4 4 5 2 0 .8 . c 0 . 79 is the estimate of the region of attraction . 5-4 Zubov’s Theorem x f ( x ) with Consider there exists V :G R h:R n n f ( 0 ) 0 and suppose R R with the following properties : (i) V is continuous ly differenti able and positive definite in G and 0 V ( x ) 1, x G { 0}. (ii) As x approaches of G , or in case of unbounded x , lim V ( x ) 1 . G as (iii) h is continuous (iv) V ( x ) Then the boundary V x and positive definite n on R . f ( x ) h ( x )[1 V ( x )], x G . x 0 is asymptotic ally stable and G is the region of attraction . 5-5 Example In the region of the origin, Hence the origin the region V ( x ) is p.d. & V ( x ) is n.d. is asymptotic ally stable. of attraction , we need to show { x G | V ( x ) 1} is invariant Ex: x1 k ( x1 ) g 2 ( x 2 ) where x 2 g 1 ( x1 ) To show and x ( 0 ) G x ( t ) 0 as t . z g ( ) d as z a or z b i i 0 i k ( 0 ) 0 , zk ( z ) 0 , a 1 z b1 g ( 0 ) 0 , zg ( z ) 0 , a z b i i i i D { x R : a i x i bi } is the region 2 G is that for some positive Show that constants a i , bi of attraction . 5-6 Example (Continued) Solution: Let h ( x ) g 1 ( x1 ) k ( x1 ). Choose V 1 W 1 ( x1 )W 2 ( x 2 ). Using W1 x1 Zubov' s theorem, V x (i.e., W 2 ( k ( x1 ) g 2 ( x 2 )) W 2 x 2 f ( x ) h ( x )[ 1 V ( x )]), W 1 ( g 1 ( x1 )) g 1 ( x1 ) k ( x1 )[ W 1 ( x1 )W 2 ( x 2 )] 0 Then W1 x1 g 1 ( x1 )W 1 ( x1 ) W 2 ( x 2 ) k ( x1 ) Then W1 x1 it is easy to Thus our choice W 2 x 2 W 1 ( x1 ) g 1 ( x 1 ) is satisfied see g 1 ( x1 )W 1 ( x1 ), W 2 x 2 by the W1 x1 W 2 ( x2 ) g 2 ( x2 ) 0 following W 1 ( x1 ), W 2 ( x 2 ) g 2 ( x 2 )W 2 ( x 2 ) of W will V 1 e 0x1 g 1 ( ) d 0x 2 g 2 ( ) d We see that V has the properties : V (0) 0, 0 V ( x ) 1, x D and V ( x ) 1 as x D 5-7 Example (Continued) And 1 g 1 ( ) d 2 g 2 ( ) d 0 V g 1 ( x1 ) ( k ( x1 ) g 2 ( x 2 )) g 2 ( x 2 )( g 1 ( x1 )) e 0 x x g 1 ( x1 ) k ( x1 )[1 V ( x )] 0 All the conditions V ( x ) is negative V 0 x1 ( t ) 0 of Zubov' s theorem semi - definite. g 1 ( x1 ) k ( x1 ) 0 g 2 ( x2 ) 0 By LaSalle' s theorem, are satisfied except th at However x1 0 x2 0 D is the region of attraction . 5-8 The most straight to use quadratic forward but conservati ve method to find Rˆ A is form (1) Linearize (2) Find Q.F. Lyapunov (3) Find the derievativ traj ectory (4) Find e of the Lyapunov of the nonlinear function along the system V 0 D where (5) Inscribe function. a level set of V in D . This Analogous procedure could be carried (1) Find V p.d. in D1 (2) Find V n.d. in D 2 is an Rˆ A . out using the direct Lyapunov method. (3) Find a level set of V in D1 D 2 This is an Rˆ A . 5-9 Advanced Stability Theory Theory for x f ( x ) is generalize d here to ( ) x f ( t , x ) ( ) x f ( t , x ) g ( t , x ) ( ) Stability if V If stable, then V (Converse Theorem) ( ) Another t ype of stability u bounded y bounded ? 5-10 Stability of time varying systems Stability of time varying systems x f ( t , x ) f : R D R (1) n where D R . n f is piecewise continuous in t and Lipschitz in x. Origin of time varying : (i) parameters change in time. (ii) investigation of stability of trajectories of time invariant system. x f ( x ) where z x x (t ) * * x ( t ) is a solution x z x (t ) * z x ( t ) f ( z x ( t )) Linearizat ioin z f ( z x ( t )) x ( t ) F ( t , z ) z A ( t ) z * * * * F z z z0 5-11 Stability • Definition of stability Definition : 0 is an equilibriu Definition : The equilibriu m of (1) if m point f (t ,0 ) 0 , 0 of (1) is stable t 0 if 0 and t 0 0 , ( , t 0 ) 0 such that x ( x0 , t0 , t ) , if The equilibriu t t0 x 0 ( , t 0 ) m point 0 of (1) is uniformly stable if 0 ( ) 0 such that x ( x0 , t0 , t ) , if t t0 x 0 ( ) 5-12 Example Ex: x ( 6 t sin t 2 t ) x dx ( 6 t sin t 2 t ) dt x ln x x t t ( 6 t sin t 2 t ) dt 0 x0 ln x ln x 0 6 sin t 6 t cos t t 6 sin t 0 6 t 0 cos t 0 t 0 2 Then x (t ) x (t0 ) e 2 2 [ 6 sin t 6 t cos t t 6 sin t 0 6 t 0 cos t 0 t 0 ] let c ( t 0 ) sup e Hence 2 2 2 [ 6 sin t 6 t cos t t 6 sin t 0 6 t 0 cos t 0 t 0 ] t t0 Then x ( t ) x ( t 0 ) c ( t 0 ), t t0 For any 0 , the choice the origin Suppose c (t0 ) shows that is stable. t 0 takes on successive and suppose x ( t ) is evaluated values t 0 2 n , n 0 ,1, 2 , seconds later in each case. 5-13 Example (Continued) Then x ( t 0 ) x ( t 0 ) exp{ 6 sin( 2 n 1) 6 ( 2 n 1) cos( 2 n 1) ( 2 n 1) 2 } 6 sin( 2 n ) 6 ( 2 n ) cos( 2 n ) ( 2 n ) } 2 x ( t 0 ) exp{ 6 ( 2 n 1) ( 2 n 1) 2 6(2n ) (2n ) } x ( t 0 ) exp{ (12 n 6 ) ( 4 n 4 n 1) 2 x ( t 0 ) exp{ 24 n 4 n 2 2 2 2 12 n 4 n } 2 2 6 } x ( t 0 ) exp{ ( 24 n 4 n 6 )} x ( t 0 ) exp{ ( 4 n 1)( 6 )} for x ( t 0 ) 0 , This implies, x (t0 ) as n x (t0 ) Thus given would 0 , there is no independen t of t 0 that satisfy t he requiremen t uniformly in t 0 . 5-14 Example (Continued) The equilibriu m point 0 of (1) is asymptotic ally stable if it is stable and 1 ( t 0 ) such that x ( x 0 , t 0 , t ) 0 as t if The equilibriu uniformly m point stable 0 of (1) is uniformly uniformly m point asymptotic The equilibriu asymptotic ally stable if it is and 1 such that x ( x 0 , t 0 , t ) 0 as t if The equilibriu x (t 0 ) 1 (t 0 ) x (t 0 ) 1 0 of (1) is globally ally stable m point uniformly asymptotic ally stable if it is and 1 . 0 of (1) is exponentia lly stable 0 , M 0 and 0 such that x ( x0 , t0 , t ) M x0 e ( t t0 ) if x0 , t t0 5-15 Example (Continued) The equilibriu exponentia m point exponentia lly stable if it is and . lly stable For time - invariant ( ) stability 0 of (1) is globally system uniformly ( ) stability There is another class of systems where the same is true – periodic system. x f ( t , x ), T 0 such that f (t T , x ) f (t , x ) Like x f (sin t , x ) Reason : it is always possible to find min ( , t 0 ) 0 t 0 [ 0 , T ) 5-16 Positive definite function • Positive definite function Class V (t , x ) K : all continuous function ( z ) such that ( 0 ) 0 and ( ) is strictly ( x ) Class K : all continuous function ( 0 ) 0 , ( ) is strictly x Definition: A continuous increasing . ( z ) such that increasing , and ( r ) as r . function V : R R R is l.p.d n if r 0 and K V (t , x ) ( x ) x r and V ( t , 0 ) 0 , t 0 if the above property radially unbounded holds and r , V is p.d and if the above holds for K . 5-17 Decrescent A continuous function V : R R R is decresent n if r 0 and K such that V (t , x ) ( x ) x r , t 0 ( x ) positive definite decrescent V (t , x ) ( x ) Thoerem: A continuous function V : R R R with V ( t , 0 ) 0 , t is n l.p.d (p.d) if and only if a l.p.d (p.d) W : R R such that n V ( t , x ) W ( x ), t , x r ( x R ) n V ( t , x ) is radially radially unbounded if this is satisfied with W ( x ) unbounded 5-18 Decrescent (Continued) V ( t , x ) is decrescent if and only if sup sup V ( t , x ) p [0, r ] x p t0 Proof : see Nonlinear systems analysis Ex: V ( t , x1 , x 2 ) ( t 1)( x12 x 22 ) x12 x 22 p.d, radially unbounded, not decrescent V ( t , x1 , x 2 ) t ( x1 x 2 ) 2 2 ( x1 x 2 )( t 1) 2 V ( t , x1 , x 2 ) 2 2 (t 2 ) 2 ( x1 x 2 )( t 1) 2 V ( t , x1 , x 2 ) not l.p.d, not decrescent 2 2 V t p.d, decrescent, radially unbounded p.d, not decrescent, not radially unbounded 2 ( x1 2 ) Finally V ( t , x ) V x f (t , x ) 5-19 Stability theorem • Stability theorem Thoerem: The equilibriu - stable m point 0 of x f ( t , x ) is if a continuous such that - uniformly ly differenti able l.p.d.f. V (t , x ) V (t , x ) 0 , t t 0 x r 0 stable if the above condition holds and V ( t , x ) if continuous ly differenti is decrescent - uniformly asymptotic l.p.d, decrescent - globally differenti uniformly ally stable V ( t , x ) such that asymptotic able p.d, decrescent V ( t , x ) such that able V ( t , x ) is l.n.d.f. ally stable and radially if a continuous ly unbounded V ( t , x ) is n.d.f. 5-20 Stability theorem (Continued) - exponentia continuous a x p lly stable if a , b , c , r 0 and p 1 and a ly differenti able V ( t , x ) such that V (t , x ) b x p , t 0, x r and V ( t , x ) c x - globally p exponentia , t 0, x r lly stable if the above Proof : same as before, holds for r ( x ) plays the role of old V ( x ). 5-21 Example Ex: y y ( 2 sin t ) y 0 x1 x 2 Mathieu eq. x 2 x 2 ( 2 sin t ) x1 2 x x V (t , x ) x 2 1 2 2 2 1 decrescent x2 2 sin t x2 ( 2 sin t ) x x2 3 positive definite 2 V 2 2 1 2 ( cos t ) 2 x1 x1 4 2 sin t cos t ( 2 sin t ) 2 2 x2 2 sin t x 2 x2 0, t , x R 2 2 0 Thus is uniformly stable. 0 5-22 Theorem Remark : LaSalle’s theorem does not work in general for time-varying system. But for periodic systems they work. So (uniformly) asymptotically stable. C onsider x ( t ) f ( t , x ) w here f ( t , x ) f ( t T , x ), x R , t 0 n Theorem Suppose V : R R n R is a continuously differentiable p.d.f and radially unbounded with V ( t , x ) V ( t T , x ), x R , t 0 n Define S { x R : V ( t , x ) 0, t 0} n Suppose V ( t , x ) 0, t 0, x R n , and that S contains no nontrivial trajectories. Under these conditions, the equilibrium point 0 is globally asymptotically stable. 5-23 Example Ex: y ( a b ( t ) cos y ) y c ( t ) sin y 0 where a 0 , b ( t T ) b ( t ) c ( t T ) c ( t ), T 0 b ( t ), c ( t ) continuous ly differenti able. b (t ) bM a , t max c ( t ) c M t min c ( t ) c m 0 t c 2 ( a b M ) c m , t Now x1 x 2 x 2 [ a b ( t ) cos x1 ] x 2 c ( t ) sin x1 5-24 Example (Continued) Choose 1 cos x1 x2 2 V ( t , x1 , x 2 ) 1 cos x1 x2 2 1 cos x1 x2 2 2cM 2 c (t ) 2cm l.p.d decrescent x2 V ( t ) 2 2 2 c (t ) Obviously 2 c ( t )[ a b ( t ) cos x1 ] c ( t ) 2 c ( t )[ a b ( t ) cos x1 ] c ( t ) 2 c m ( a b M ) c and 2 c m ( a b M ) c 0 , t So, V 0 Now we again use the invariance principle V 0 x 2 0 x1 const. c ( t ) sin x1 0 x1 0 (Uniformly ) asymptotic ally stable 5-25 Instability Theorem (Chetaev) • Instability Theorem (Chetaev) The equilibriu differenti m point 0 of x f ( t , x ) is unstable ly able V ( t , x ), set B r { x R : x r }, an open set B r and a function n r K such that 0 V ( t , x ), t 0 , x sup sup V ( t , x ) t if a continuous x 0 V (t , x ) 0 , t , x B r V 0 V 0 V 0 V 0 V ( t , x ) r ( x ), t , x 5-26 Linear time-varying systems and linearization Linear time-varying systems and linearization x A ( t ) x x R n x (t ) (t , t0 )x (t0 ) state transitio n matrix d dt (t , t0 ) A (t ) (t , t0 ) (t0 , t0 ) I , Stability analyze t0 0 of linear ti me varying as that of nonlinear system is almost as difficult to system. 5-27 Example Ex: x A ( t ) x 1 1 . 5 cos 2 t A (t ) 1 1 . 5 sin t cos t 1 1 . 5 sin t cos t 2 1 1 . 5 sin t det( I A ( t )) 0 1 , 2 0 . 25 j 0 . 25 7 e 0 . 5 t cos t But ( t , 0 ) 0 .5 t sin t e looks like stable?? e e t t sin t cos t Thus x 0 such that x ( x 0 , t 0 , t ) as t 5-28 Theorem Theorem: The equibrium 0 of x A ( t ) x is stable point iff sup ( t , t 0 ) , t 0 t t0 - uniformly stable iff sup sup ( t , t 0 ) t0 0 - (globally) t t0 uniformly (t , t 0 ) - (globally) i i asymptotic ally stable iff the above condition holds 0 as t , t 0 uniformly asymptotic ally stable iff r (t t ) 0 ( t , t 0 ) i Ke , t t0 0, K 0, r 0 exponentia lly stable Proof : See Nonlinear systems analysis 5-29 Lyapunov function approach • Lyapunov function approach x A ( t ) x V ( t , x ) x P ( t ) x where T Then 2 c1 I P ( t ) c 2 I , t , c i 0 2 c1 x x P ( t ) x c 2 x T p.d decrescent T T T V ( t , x ) x P ( t ) x x P ( t ) x x P ( t ) x T T T T x A ( t ) P ( t ) x x P ( t ) x x P ( t ) A ( t ) x x T P ( t ) A x (t ) x T i.e. V ( t , x ) c 3 x T (t ) P (t ) P (t ) A (t ) x where (t ) c3 I , t , c3 0 2 Result : exponentia lly stability Note that P ( t ) is defined by a positive definite symmetric solution of T P ( t ) P ( t ) A ( t ) A ( t ) P ( t ) ( t ) 5-30 Theorem Theorem: x f ( t , x ), f : R R n R n continuous ly diff. f (t ,0 ) 0 , t 0 lim sup f1 ( t , x ) x 0 t0 A (t ) 0 where x f (t , x ) x is bounded f1 ( t , x ) f ( t , x ) A ( t ) x t x0 Then 0 is an exponentia lly stable eq. point of x f ( t , x ) if it is exponentia lly stable for x A ( t ) x . Proof : See Nonlinear systems analysis 5-31 Converse (Inverse) Theorem & Invariance Theorem Converse (Inverse) Theorem • i) if V stable • ii) (uniformly asymptotically exponentially) stable V Invariance Theorem : positive V : R , V 0 in E { x : V 0} M : largest invariant set in E x ( x 0 , t 0 , t ) M as t , x 0 invariant We can eliminate set indirect how to define E is not clear since V is a function of t , x as well. x D set E { x D : W ( x ) 0} can be defined analogous case, the uncertaint y by assuming V ( t , x ) W ( x ) 0 Then the In time varying of the LaSalle' s theorem as before. Thus an can be formed as follows : 5-32 Theorem Theorem : Let D { x R : x r } where 2 and locally function Lipschitz x f ( t , x ) is piecewise in X , uniformly continuous in t in t . Let V be a cont. diff. such that W1 ( x ) V (t , x ) W 2 ( x ) V V V ( t , x ) f ( t , x ) W ( x ), t 0 , x D t x where W 1 ( ), W 2 ( ) are continuous a continuous positive Then all solutions bounded positive semidefini te function of x f ( t , x ) with definite functions and W ( ) is on D . Let min W 1 ( x ). x r x ( t 0 ) { x B r : W 2 ( x ) } are and W ( x ( t )) 0 as t Proof : See Ch 4.3 of Nonlinear Systems x ( t ) approaches Therefore E as t since W ( x ( t )) 0 as t . the positive limit set of x ( t ) is a subset of E . 5-33