The smallest upper bound on the number of zeros of abelian integrals (Q2180598)

The authors provide an algebraic criterion to study the maximum number of zeroes of a linear combination of abelian integrals. This algebraic criterion avoids the explicit computation of the abelian integrals and considers the zeroes in their integrand. It can be applied to show that the set of abelian integrals forms an ECT-system with accuracy zero or higher. The criterion can be used in a (slightly) more general setting, but the framework is a Hamiltonian system with first integral of the form \[ H(x,y) = \frac{y^2}{2} + \Psi(x) \] where \(\Psi(x)\) is an analytic an function in an open interval \((\mu,\nu)\), which corresponds to the maximal open interval on which the oval \(\Gamma_h\) the energy level set \(\{(x, y) : y^2/2+\Psi(x) = h \}\) exists. Then the authors consider a small perturbation of this system such that the bifurcation function at first order of the perturbation parameter can be written as \(I(h) = \sum_{i=1}^{k} \alpha_i I_i(h)\) with \(\alpha_i\) real constants depending on the parameters of the perturbation and \[ I_i(h) = \oint_{\Gamma_h} f_i(x) y \, dx, \] with \(f_i(x)\) analytic functions on \((\mu,\nu)\), \(i=1,2,\ldots,k\). The criterion consists in defining a set of analytic functions depending on \(x\) which come from \(\Psi(x)\) and \(f_i(x)\) and studying its zeroes. Then, the criterion is used for determining the smallest upper bound of the number of isolated zeros of an abelian integral for several systems, as for instance one whose Hamiltonian is of the form \[ \left( y^2 + \frac{1}{2} \right) \frac{1}{x^2} + \ln |x| - \frac{1}{2} = h, \quad \text{with } 0<h<\infty. \]