Cyclicity of a class of polynomial nilpotent center singularities (Q256217)

The authors consider planar analytic differential systems of the form \(\dot{x} \, = \, f(x)\) where the origin is an isolated monodromic singularity. The stability of the origin is the leitmotif of the paper. When \(df(0)\) is nondegenerate, the origin is monodromic only if its eigenvalues have nonzero imaginary part. In such a case, the classical theory due to Poincaré and Liapunov establishes that a set of Poincaré-Liapunov quantities \(v_i\), which are polynomials on the coefficients of the system, determine whether the origin is a stable focus, a unstable focus or a center. The Poincaré-Liapunov quantities \(v_i\) are the coefficients of the development of displacement map in a neighborhood of the origin. The stability of the origin can also be studied using the focus quantities \(f_i\), which are the coefficients of the temporal derivative of a Liapunov function. In the nondegenerate case it is well known that the two sets, namely the Poincaré-Liapunov quantities and the focus quantities, provide the same information about the stability of the origin. In the nilpotent case, that is, when \(df(0)\) is a nilpotent matrix which is not identically null, then the celebrated Andreev Theorem characterizes when the origin is monodromic. The authors focus their attention to systems with a nilpotent monodromic point at the origin. First, the authors characterize when the Poincaré-Liapunov quantities \(v_i\) are polynomials in the coefficients of the system. This sentence, which is loosely used for nondegenerate systems, is clarified in this work. The computation of the Poincaré-Liapunov quantities \(v_i\) for a nilpotent singular point is done, after some changes of variables, by means of a change to generalized polar coordinates \(x = r \, \text{Cs} \, \theta\), \(y = r \, \text{Sn}\, \theta\). The system is transformed to an ordinary differential equation of the form \[ \frac{dr}{d\theta} \, = \, \mathcal{F}(r,\theta), \] where \(\mathcal{F}(r,\theta)\) is analytic in a neighborhood of \(r = 0\), is \(T_n\)-periodic in \(\theta\), and satisfies \(\mathcal{F}(0,\theta) \equiv 0\). We recall that \(T_n\) is the period of the generalized trigonometric functions \(\text{Cs}\, \theta\) and \(\text{Sn} \, \theta\). With this equation, one can compute an expression of the displacement map in a neighborhood of the nilpotent singularity, and thus, the Poincaré-Liapunov quantities. Note that these quantities are very difficult to be computed as the integration of generalized trigonometric polynomials on its period needs to be done. Then the authors consider systems of the form \[ \dot{x} \, = \, y + P_{2m+1}(x,y), \quad \dot{y} \, = \, Q_{2m+1}(x,y), \] where \(P_{2m+1}\) and \(Q_{2m+1}\) are homogenoeus polynomials of degree \(2m+1\). The authors show how to compute the focal quantities for the nilpotent monodromic singular point at the origin of the former system. Note that this computation relies on solving linear systems of equations and are far easier to be computed than the Poincaré-Liapunov quantities. One of the main results of the paper is that there is a relation between the ideals generated by \(v_1, v_2, \ldots, v_k\) and \(f_1, f_2, \ldots, f_k\), with \(k \in \mathbb{N}\), where \(v_i\) denote the Poincaré-Liapunov quantities and \(f_i\) denote the focus quantities associated to the nilpotent monodromic point at the origin of the previous system with homogeneous nonlinearities. In this way, the cyclicity of the origin of this system can be handled with the same tools as in the nondegenerate case. The authors study the cyclicity of the origin of these systems in the cubic (\(m=1\)) and the quintic (\(m=2\)) cases. They establish a sharp upper bound of the cyclicity for the cubic case and also for a broad subclass of the systems in the quintic case. We remark that the study of the cyclicity is far more involved than the study of the focus quantities. The focus quantities provide a previous step in the study of the cyclicity but many concepts from computational algebra and the study of ideals of polynomials arise when tackling the cyclicity problem.