next up previous contents
Next: Hyperbolic Fixed Points: . Up: Fixed Points Previous: Stability

Linearization

   To calculate the stability of a fixed point consider a small perturbation, tex2html_wrap_inline15114 , about tex2html_wrap_inline15098 ,

 

The Taylor expansion  (substituting eq. (4.16) into eq. (4.2)) about tex2html_wrap_inline15098 gives

  equation4858

It seems reasonable that the motion near the fixed point should be governed by the linear system

  equation4872

since tex2html_wrap_inline15120 . If tex2html_wrap_inline15122 is an equilibrium point, then tex2html_wrap_inline15124 is a matrix with constant entries. We can immediately write down the solution to this linear system as

  equation4888

where tex2html_wrap_inline15126 is the evolution operator for a linear system. If we let tex2html_wrap_inline15128 denote the constant tex2html_wrap_inline14512 matrix, then the linear evolution operator  takes the form

  equation4903

where tex2html_wrap_inline15132 denotes the tex2html_wrap_inline14512 identity matrix.

The asymptotic stability of a fixed point can be determined by the eigenvalues of the linearized vector field tex2html_wrap_inline15136 at tex2html_wrap_inline15098 . In particular, we have the following test for asymptotic stability: an equilibrium solution of a nonlinear vector field is asymptotically stable if all the eigenvalues of the linearized vector field tex2html_wrap_inline15124 have negative real parts.

If the real part of at least one eigenvalue exactly equals zero (and all the others are strictly less than zero) then the system is still linearly stable, but the original nonlinear system may or may not be stable.



Nicholas B. Tufillaro
Mon Mar 3 01:58:02 PST 1997