scientific article; zbMATH DE number 3847813

From MaRDI portal
Publication:3317404





The problem of the boundedness of the solutions of the differential equation \[ (1)\quad \ddot x+kf(x,\dot x,t)\dot x^{2n-1}+g(x)=\epsilon h(x,\dot x,t)\quad(\dot x=dx/dt) \] of order 2 in the real domain is reduced to the corresponding problem for the differential inequality \[ (2)\quad {\dot \eta}+\lambda \eta +g(x)sgn \dot x\leq \mu, \] where \(\eta =| \dot x|\), and \(\lambda =\lambda(\epsilon /k)>0\), \(\mu =\mu(\epsilon /k)>0\) are constants, by the following theorem: T.1. Every solution of the differential equation (1) satisfies the differential inequality (2). One of the general deductions made from T.1 is T.2. If (i) \(ug(u)>0\) when \(| g(u)| \geq \mu\), (ii) \(\nu>0\) exists such that \(| g(u)| \geq \mu\) when \(| u| \geq \nu\), and such that \(| g(u)|<\mu\) when \(| u|<\nu\); (iii) \(| g(u)| \leq \mu\) implies \(| u| \leq \mu '\), a constant. Then every solution of (1) satisfies \(x^ 2(t)+x^ 2(t)=0(1)\) as \(t\to \infty\). The basic hypotheses on (1) are that \(k>0\), \(\epsilon>0\) are parameters; \(n\geq 1\) is a rational with a positive odd denominator; \[ | f(u,\nu,t)| \geq 1;\quad | h(u,\nu,t)| \leq 1+\nu^{2n-2};\quad f(u,\nu,t),g(u),h(u,\nu,t) \] are continuous. Some boundedness theorems are established for the special case \[ (3)\quad \ddot x+k\Phi(\dot x)f(x)\dot x+g(x)=\epsilon \Phi(\dot x)h(t) \] of equation (1) and generalisations of theorems 1 and 2 to cover the generalisation ẍ\({}_ r+k_ rf_ r\dot x_ r^{2nr-1}+g_ r=\epsilon h_ r (r=1,...,s)\) of equation (1) are obtained.











This page was built for publication:

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3317404)