No label defined (Q4337797)

In 1984 \textit{P. J. Ponzo} and \textit{N. Wax} proved an existence theorem of limit cycles for the system: \[ \dot x=y,\;\dot y=-f(x,y) y-g(x) \tag{1} \] [see J. Math. Anal. Appl. 104, 117-127 (1984; Zbl 0618.34034)]. Then \textit{Z. H. Zheng} [J. Math. Anal. Appl. 148, 1-10 (1990; Zbl 0706.34037)] gave a counter-example: \[ \dot x=y,\;\dot y=-(x^2-4) (y^2+1)y-x \tag{2} \] by showing that (2) has no limit cycle, and also improved the theorem of Ponzo and Wax. In 1994 \textit{J. S. Yu} and \textit{L. H. Huang} [Differ. Integral Equations 7, No. 2, 501-507 (1994; Zbl 0790.34041)] generalized the theorem of Ponzo and Wax to the system \[ \dot x=y,\;\dot y=-f(x,y) \varphi(y) -g(x) \tag{3} \] and showed that (2) has at least one limit cycle. In this paper the authors prove that Zheng's conclusion is correct but Ponzo and Wax's as well as Yu and Huang's theorems are incorrect. They also give some theorems on the boundedness of all positive semi-orbits of (3) as well as a theorem on the nonexistence of limit cycles of (3). These are corrections of theorems by Ponzo and Wax and Huang and Yu, and an improvement of a theorem of Zheng.