No label defined (Q4358917)

By using the transformation \(w= {1\over z^2}\) to change the Abel equation: \[ {dz\over d\theta} =A(\theta)z^3 +B(\theta) z^2+C(0)z \tag{1} \] (where \(A(\theta + \pi) =A(\theta)\), \(B(\theta +\pi)= \pm B(\theta)\), \(C(\theta +\pi) =C(\theta))\) into: \[ {dw\over d \theta} =-2C(\theta) w-2A (\theta)\pm 2B(\theta) \sqrt w\quad (z\lessgtr 0) \tag{2} \] the author gets some theorems on the non-existence, existence or uniqueness of limit cycles of (1). In (1) the \(A(\theta)\) and \(B(\theta)\) may now change their sign as \(\theta\) varies. One of the theorems is used to prove that the cubic system: \[ \dot x=x- y-x^3- \sqrt 6x^2y-xy^2, \quad \dot y=x +y- x^2y- \sqrt 6xy^2 -y^3 \] has a unique limit cycle.