I am trying to understand a bit about solutions of delay differential equations, so I tried analyzing one of the most simple ones: $$u'(t)=-\beta u(t-1), \text{and for } t\in [-1,0), u(t)=\phi(t), \text{for some known function $\phi$}.$$

One way to obtain solutions is to form the characteristic equation by substituting $u(t)=e^{\lambda t}$. Using this method, one can prove that if $\beta=e^{-1}$, then $e^{-t}$ and $te^{-t}$ satisfy $u'(t)=-\beta u(t-1)$. These solutions don't take into account the initial function $\phi$.

Another way of obtaining a solution is to start with the function $\phi$ and integrate the equation in every interval of length $1$. This method is called method of steps.

My question is, what is the relationship between the solutions one obtains from the characteristic equation method and the method of steps?

  • $\begingroup$ If a solution does not take into account a part of the problem (initial condition), then it is not a solution. $\endgroup$ – user53153 Jan 23 '13 at 14:32

The general solution is $u(t) = \sum_n c_n \exp(\lambda_n t)$ where $\lambda_n $ are the solutions of the characteristic equation. The $c_n$ are obtained from $\phi$ using a Laplace transform, which is explained for example in here, but I have not read it in detail so I do not know if the relation is simple, https://www.rand.org/content/dam/rand/pubs/reports/2006/R374.pdf

This solution must be equal to the one obtained with the method of steps with the same initial condition.


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.