2
$\begingroup$

I am applying at a Faculty for Electrical Engineering, and I have an entrance exam in two days. I have few exams for exercise, from previous years and I keep getting stuck in one particular type of problem:

"A dynamicaI system is described by a differential equation of the following form: $$y''(t) + 5y'(t)+4y(t) = u(t)$$ where t is the independent variable of time, $u(t)$ is the input signal into the system, and $y(t)$ is the system response."

I need to prove that if the system is excited with stepped amplitude signal of $1$, the output of the system in the steady state will reach a constant value of $0.25$.

Can anyone please suggest how can I solve this problem. I think I should use Laplace Transformation to solve it, but I do not know to how transform the above differential equation in some type that can be solved with Laplace. Also, i do not get where does the amplitude signal of $1$ is used.

Any help or hints are welcomed and appreciated, I need all the help I can get. Thank you.

$\endgroup$
  • $\begingroup$ Maybe this would suit ELECTRICAL ENGINEERING.SE better? $\endgroup$ – Mr. Xcoder Aug 29 '17 at 9:21
  • $\begingroup$ Thank you for the suggestion, I did not know there is a page for Electrical Engineering. I will try there. $\endgroup$ – mchingoska Aug 29 '17 at 9:33
  • $\begingroup$ The question definitely belongs to math.SE. To solve this, why not use that every solution of the differential equation is $$y(t)=ae^{-t}+be^{-4t}+\frac13e^{-t}\int_0^tu(s)e^sds-\frac13e^{-4t}\int_0^tu(s)e^{4s}ds$$ for some constants $(a,b)$ depending on the initial conditions $(y(0),y'(0))$, and plug in the function $u(t)$ you are interested in, whatever it is? For instance, if $u(t)=1$ for every $t>0$, one gets a fully explicit formula for $y(t)$, from which the limit when $t\to\infty$ is direct... $\endgroup$ – Did Aug 29 '17 at 9:36
  • $\begingroup$ An electrical engineer would probably do Laplace tranformation $$(s^2+5s+4)Y=\frac{1}{s}$$ then do the partial fractions, solve the system and ensure that the limit is as expected. Or maybe use the final value theorem instead if allowed. $\endgroup$ – A.Γ. Aug 29 '17 at 10:05
  • $\begingroup$ I started solving it like that and ended with $$Y=\frac{1}{4}-\frac{1}{3}e^{-t}+\frac{1}{12}e^{-4t}$$ , and I am stuck again. $\endgroup$ – mchingoska Aug 29 '17 at 10:17
0
$\begingroup$

To answer your question you do not actually need to solve the differential equation. Indeed, in steady state conditions both $y´´$ and $y´$ will vanish: hence you are (asymptotically) left with the equation $$ 4 y(t) = 1$$ (as $u(t)$ is where the step input function comes into play) from which one concludes $$ y = 1/4 = 0.25$$

| cite | improve this answer | |
$\endgroup$
  • 1
    $\begingroup$ Assuming there is a steady state... $\endgroup$ – Did Aug 29 '17 at 9:40
  • $\begingroup$ @Did, needless to say you are right. I was just wondering if the existence of a steady state was in some way implied by the question formulation itself, which I believe the OP adapted from the exam sheet, "I need to prove that if the system is excited with stepped amplitude signal of 1 , the output of the system in the steady state will reach a constant value of 0.25" $\endgroup$ – An aedonist Aug 29 '17 at 9:42
  • $\begingroup$ Yes, the steady state was given in the question formulation. I did not know that in steady state the condition y'' and y' vanish. @An aedonist, your answer helped me a lot, thank you very much. $\endgroup$ – mchingoska Aug 29 '17 at 9:51
  • 1
    $\begingroup$ @KirylPesotski If you feed in the harmonic signal $e^{i\omega t}$ you will get your output $y_p(t)$ as a harmonic oscillation with no existing limit as $t\to\infty$. Your $y_s$ is not a steady state, but an amplitude (frequency response at $\omega$). I do not see your point. $\endgroup$ – A.Γ. Aug 29 '17 at 11:01
  • 1
    $\begingroup$ in the harmonic case, the steady state solution does not exist! $\endgroup$ – Kiryl Pesotski Aug 29 '17 at 11:19

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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