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.

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

Conside the differential equation $$\dot{x}=Ax,\qquad x(t):{\bf R}\to{\mathcal H}$$ where $\mathcal{H}$ is a Hilbert space and $A$ is a bounded linear operator. With the initial condition, one can have $x(t)=e^{At}x_0$ (Is this legal in the infinite dimension case?). With the spectral method(under the assumption that $x_0$ is an eigenvector of $A$), one has the estimate $$\|x(t)\|\leq e^{\omega t}\|x_0\|.$$

Here is my question:

  • With some additional assumption, can one estimate $\|x(t)\|$ without using the spectral method, say, simply taking the inner product?

I think the following sub-question may be helpful:

  • What is $\frac{d}{dt}\langle x(t),x(t)\rangle$?

  • In the finite dimension case, this can be done with the product rule. What about the infinite dimension case?

share|cite|improve this question
up vote 9 down vote accepted

Yes, if $A$ is a bounded linear operator $e^{At}$ is quite OK. It can be defined by the Taylor series, for example, and [EDIT: for $t \ge 0$ ] we have $$\|e^{At}\| \le e^{\|A\| t}.$$ If $x_0$ is an eigenvector for $A$ with eigenvalue $\lambda$, then $x(t) = e^{\lambda t} x_0$.

As for your sub-question, it is still true that $$\frac{d}{dt} \langle x(t),\, x(t) \rangle = \langle x'(t), x(t) \rangle + \langle x(t), x'(t) \rangle = 2 \Re \langle x(t), A x(t) \rangle.$$ The proof using difference quotients is basically the same as in single-variable calculus.

share|cite|improve this answer

Your Answer


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.