No label defined (Q4026084)

From MaRDI portal

!

This is the item page for this Wikibase entity, intended for internal use and editing purposes. Please use the normal view instead:

scientific article; zbMATH DE number 124458
Language Label Description Also known as
default for all languages
No label defined
    English
    No label defined
    scientific article; zbMATH DE number 124458

      Statements

      21 February 1993
      0 references
      Hilbert's sixteenth problem
      0 references
      number of limit cycles
      0 references
      polynomial vector field in a plane
      0 references
      0 references
      [For the entire collection see Zbl 0745.00036.] The second half of Hilbert's sixteenth problem is to find a bound on the number of limit cycles that can exist for a polynomial vector field in a plane. An obvious extension of this problem is to find a bound for polynomial vector field in \(R^ n\). This is the concern of the paper which is to explore the idea of obtaining such a bound by bounding the number of solutions to a related discrete problem. More specifically, the paper considers the system of differential equations \(du/dt=TF(u)\), \(u(0)-u(1)=0\); \(dv/dt=u(t)\cdot V^ 0(t)\), \(v(0)=1\); \(dT/dt=0\), \(v(1)=1\); or the more simply written equations \[ G(y)\equiv\begin{cases} dy/dt- f(t,y) & =0\\ g(y(0),y(1)) & =0\end{cases}\tag{1} \] where \[ y(t)=\left[\begin{matrix} u(t)\\ v(t)\\ T(t)\end{matrix}\right],\quad f(t,y)=\left[\begin{matrix} TF(u)\\ u\cdot V^ 0(t)\\ 0\end{matrix}\right],\quad g(y(0),y(1))=\left[\begin{matrix} u(0)-u(1)\\ v(0)\\v(1)\end{matrix}\right]; \] here, \(F\), \(u\), \(V^ 0\) are \(n\)-vectors, \(v\), \(T\) are scalars, \(T\) being the unknown period, and \(V^ 0(t)\) may be reasonably taken as \(F^ 0(t)=\sin(2\pi t)e^ T\), \(e=[1,1,\dots,1]\). A discretization of system (1) leads to the equation \[ G^ h(y^ h)\equiv\begin{cases} (y_ j-y_{j- 1})/h-f(t_{j-\textstyle{1\over 2}},(y_ j+y_{j-1})/2) & =0,\;1\leq j\leq J;\\ g(y_ 0,y_ J) & =0\end{cases}\tag{2} \] Here, \(y_ j\in R^{n+2}\) is to approximate \(y(t_ j)\), \(t_ j=jh\), for \(j=0,1,\dots,J\); \(h=1/J\). Under appropriate assumptions, the author showed that isolated solutions of (1) imply isolated roots of (2). In fact, the following theorem is obtained. Theorem. Let \(G^ h(\cdot)\) as in (2) and \(G(\cdot)\) as in (1) satisfy: (i) \(G^ h(\cdot)\) is accurate of order \(p\) for \(G(\cdot)\); (ii) The linearization \(L^ h(x^ h)\equiv G^ h_{x^ h}(x^ h)\) are Lipschitz continuous with bounded inverses. Then for \(h\leq h_ 0\) sufficiently small if (1) has an isolated solution, \(y(t)\), it follows that (2) has an isolated solution, \(y^ h=[y_ 0,y_ 1,\dots,y_ J]\), and \(\| y_ j-y(t_ j)\|\leq K_ 0 M(y)h^ P\), \(j=0,1,\dots,J\). \(K_ 0\), \(M(y)\) are constants. --- The paper concludes with a discussion on the difficulties in applying the results to the Hilbert problem for all polynomial vector fields.
      0 references

      Identifiers