Take the 2-minute tour ×
Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. It's 100% free, no registration required.

Let vector-function $x(t)$ satisfy a differential equation $$ \dot x = f(x), $$ and a vector-function $y($t) satisfy a differential inequality $$ \dot y \leq f(y) $$ with starting positions $y(0) < x(0)$. If a function $f(x)$ satisfies the property: $$ f_{i}(x_1+\alpha_1,\ldots,x_{i-1}+\alpha_{i-1},x_i,x_{i+1}+\alpha_{i+1},\ldots,x_{n}+\alpha_{n}) \geq f_{i}(x_1,\ldots,x_n) $$ for any $\alpha_{1} \geq 0, \ldots, \alpha_{n} \geq 0$ (i.e. it is quasimonotone), then $y(t) \leq x(t)$ for any $t>0$. Function $f(x)$ is smooth.

Is there a name for such theorem? Please help me to proof it or give me a reference.

share|improve this question

2 Answers 2

up vote 1 down vote accepted

This theorem is known in Russia as the Chaplygin lemma. It can be proved as follows. Suppose that it isn't true. Then let $$ t^{*} = \inf \{ t \geqslant 0 \mid \exists i\colon y_{i}(t) > x_{i}(t) \} < \infty $$ By definition of $t^{*}$ we have that $y_{i}(t^{*}) = x_{i}(t^{*})$ and for any $j \neq i$ we have $y_{j}(t^{*}) \leqslant x_{j}(t^{*})$. Then by the quasimonotony propertie we have $$ f_{i}(y(t^{*})) \leqslant f_{i}(x(t^{*})) \tag{0} $$ On the other hand by the definition of $t^{*}$ there exists some small $\delta > 0$ such that $$ y_{i}(t^{*}+\Delta t) > x_{i}(t^{*} + \Delta t) \tag{1} $$ for any $0 < \Delta t < \delta$. Then $$ \dot y_{i}(t^{*}) \geqslant \dot x_{i}(t^{*}) = f_{i}(x(t^{*})) \tag{2} $$ because the opposite inequality implies contradiction with $(1)$. There may occur two different situations.

  1. $\dot y_{i}(t^{*}) < f_{i}(y(t^{*}))$. From $(2)$ it follows that $f_{i}(y(t^{*})) > f_{i}(x(t^{*}))$. This is a contradiction with $(0)$.
  2. $\dot y_{i}(t^{*}) = f_{i}(y(t^{*}))$. Then consider a solution $y_{\varepsilon}(t)$ of differential inequality $$ \dot y_{\varepsilon} \leqslant f(y_{\varepsilon})-\varepsilon $$ From the first case we have that $y_{\varepsilon}(t) \leqslant x(t)$ for any $t$. Then let $\varepsilon \to 0^{+}$ and use that the solution depends continuously of parameter $\varepsilon$.

Theorem is proved.

share|improve this answer

W. Walter, Ordinary Differential Equations, Springer (1998)

See section 9, page 89 -- 97 .

share|improve this answer

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.