- The fixed point is asymptotically stable if it is Lyapunov stable and there is a neighbourhood of such that as for every .
- Its domain of stability, or basin of attraction, is , with forward existence understood.
- A Lyapunov function on a neighbourhood of is a continuous function , differentiable away from the origin, such that , for , and its orbital derivative
The First Lyapunov theorem says that the existence of such a function makes the origin Lyapunov stable. To prove it, choose a sufficiently small closed ball . By compactness and positive definiteness,By continuity, some satisfies whenever . Along a trajectory beginning in , cannot increase. Such a trajectory therefore cannot reach the sphere , where . Hence it remains in for every , which is Lyapunov stability.
The LaSalle invariance principle says that if a trajectory remains in a compact positively invariant set on which , then it approaches the largest invariant subset of .
On the two vertical sides of the square, the component of the vector field satisfiesOn the horizontal sides,Thus the vector field never points outward through the boundary, so the closed square is a positively invariant set and trajectories cannot leave .
At a fixed point, gives . Substitution into givesFor , both and are positive; for , both are negative. Hence , and then , is the unique equilibrium point in .
This alone does not prove that lies in the domain of stability. A compact trapping region containing a unique equilibrium can also contain a periodic orbit or another nontrivial omega-limit set.
Forthe orbital derivative isChoosing cancels the mixed term. Taking givessoThis function has positive definiteness near the origin, andFor and , both terms are nonpositive and equality holds only at . Thus is a strict Lyapunov function near the origin, and the Second Lyapunov theorem proves that the origin is asymptotically stable.
The factor is positive for . The product is positive for . Therefore away from the origin throughoutNeither side length can be enlarged while retaining a strict Lyapunov function on the whole rectangle: at with , the first term makes , while at immediately beyond , the second term makes .
Strict decrease on this rectangle does not by itself put the whole rectangle in the domain of stability, because the rectangle has not been shown to be positively invariant; a trajectory may leave it before convergence.
To obtain a basin estimate from a Lyapunov sublevel set, take the largest sublevel component around the origin that remains inside the strict-decrease rectangle. On its boundary or , the minimum value of is . Hence the maximal open estimate isEvery smaller closed sublevel set is compact and positively invariant, and there except at the origin. The LaSalle invariance principle therefore implies convergence to the origin. Taking the union over shows that lies in the domain of stability. No larger sublevel component can be certified from this alone, because every level contains points just beyond where .
Articles by others on the same topic
There are currently no matching articles.