Linear first-order differential equation with an initial condition

Click For Summary
The discussion focuses on solving a linear first-order differential equation with initial conditions, specifically addressing challenges with derivatives and integrals. The user is attempting to derive an expression for the function f(x) and is struggling particularly with the second term involving an integral and its derivative. They explore different approaches to simplify the equation, including removing integrals from exponential terms. The conversation highlights the application of the product rule and the evaluation of integrals at specific points to clarify the calculations. Overall, the thread emphasizes the complexities involved in differentiating expressions that include integrals and exponential functions.
Lambda96
Messages
233
Reaction score
77
Homework Statement
Show that the solution of a linear differential equation with boundary condition ##f(0)=f_0## has the following form ##f(x)=f_0 exp\biggl( \int_{0}^{x}ds ~g(s) \biggr)+\int_{0}^{x}ds ~h(s)exp\biggl( \int_{s}^{x}dr ~g(r)\biggr)##
Relevant Equations
none
Hi,

unfortunately I have problems with the task d and e, the complete task is as follows:

Bildschirmfoto 2023-06-28 um 14.05.38.png

Bildschirmfoto 2023-06-28 um 14.06.04.png

I tried to form the derivative of the equation ##f(x)##, but unfortunately I have problems with the second part, which is why I only got the following.

$$\frac{d f(x)}{dx}=f_0 g(x) \ exp\biggl( \int_{0}^{x}ds \ g(s) \biggr)+?$$

I then tried another approach and first wanted to get rid of the integrals in the exp terms

$$f_0 exp\biggl( G(x)-G(0) \biggr)+\int_{0}^{x}ds \ h(s) exp \biggl( G(x)-G(s) \biggr) f(x)dx$$
$$f_0 e^{G(x)-G(0)}+e^{G(x)}\int_{0}^{x}ds \ h(s) e^{-G(s)}$$

Unfortunately, I now have problems again with the derivative for the second term using this method.
 
Physics news on Phys.org
You have an expression of the form <br /> \int_a^x f(x,t)\,dt for which you can show from the limit definition that, for sufficiently smooth f, <br /> \frac{d}{dx} \int_a^x f(x,t)\,dt = \lim_{h \to 0} \frac 1h \left( \int_a^{x+h} f(x+h, t)\,dt - \int_a^x f(x,t)\,dt\right) =<br /> f(x,x) + \int_a^x \left.\frac{\partial f}{\partial x}\right|_{(x,t)}\,dt.
 
  • Like
Likes DeBangis21 and Lambda96
Thanks pasmith for your help 👍 , I tried to implement your tip but am not sure about the ##f(x,x)## term:

I have now calculated the following

$$\frac{d f(x)}{dx}=\frac{d}{dx}\biggl( f_0 \ exp\Bigl( \int_{0}^{x}ds \ g(s) \Bigr) +\int_{0}^{x}ds \ h(s) \ exp\Bigl( \int_{s}^{x}dr \ g(r) \Bigr) \biggr)$$

$$\frac{d f(x)}{dx}= f_0 \ exp\Bigl( \int_{0}^{x}ds \ g(s) \Bigr) g(x) + h(x)exp\Bigl( \int_{s}^{x}dr \ g(r) \Bigr)+\int_{0}^{x}ds \ h(s) \frac{d}{dx} \ exp\Bigl( \int_{s}^{x}dr \ g(r) \Bigr)$$

$$\frac{d f(x)}{dx}= f_0 \ exp\Bigl( \int_{0}^{x}ds \ g(s) \Bigr) g(x) + h(x)exp\Bigl( \int_{s}^{x}dr \ g(r) \Bigr)+\int_{0}^{x}ds \ h(s) exp\Bigl( \int_{s}^{x}dr \ g(r) \Bigr)g(x)$$

$$\frac{d f(x)}{dx}= f_0 \ exp\Bigl( \int_{0}^{x}ds \ g(s) \Bigr) g(x) + h(x)exp\Bigl( \int_{s}^{x}dr \ g(r) \Bigr)+\int_{0}^{x}ds \ h(s) exp\Bigl( \int_{s}^{x}dr \ g(r) \Bigr)g(x)$$

$$\frac{d f(x)}{dx}= f_0 \ exp\Bigl( \int_{0}^{x}ds \ g(s) \Bigr) g(x) +\int_{0}^{x}ds \ h(s) exp\Bigl( \int_{s}^{x}dr \ g(r) \Bigr)g(x)+ h(x)exp\Bigl( \int_{s}^{x}dr \ g(r) \Bigr)$$

$$\frac{d f(x)}{dx}=f(x)g(x)+ h(x)exp\Bigl( \int_{s}^{x}dr \ g(r) \Bigr)$$

Now I'm just having trouble with the term ##h(x)exp\Bigl( \int_{s}^{x}dr \ g(r) \Bigr)##, which is equivalent to ##f(x,x)##. I was just having trouble figuring out what the exp term should be. After all, the easiest way would be for the integral to go from x to x, which would make the exponent zero and thus ##e^{0}=1## and I would get ##f(x)g(x)+h(x)##
 
Lambda96 said:
Now I'm just having trouble with the term ##h(x)exp\Bigl( \int_{s}^{x}dr \ g(r) \Bigr)##, which is equivalent to ##f(x,x)##. I was just having trouble figuring out what the exp term should be. After all, the easiest way would be for the integral to go from x to x, which would make the exponent zero and thus ##e^{0}=1## and I would get ##f(x)g(x)+h(x)##

Yes. The integrand is a function of x and s, so it is evaluated at (x,s) = (x,x).
 
Lambda96 said:
I then tried another approach and first wanted to get rid of the integrals in the exp terms

$$f_0 \exp\biggl( G(x)-G(0) \biggr)+\int_{0}^{x}ds \ h(s) \exp \biggl( G(x)-G(s) \biggr) \,dx$$
$$f_0 e^{G(x)-G(0)} + e^{G(x)} \int_{0}^{x}ds \ h(s) e^{-G(s)}$$

Unfortunately, I now have problems again with the derivative for the second term using this method.
The second term is of the form ##e^{G(x)} [J(x)-J(0)]## where ##J'(x) = h(x) \exp[-G(x)]##. Just apply the product rule.
 
Thanks pasmith and vela for your help 👍👍
 
First, I tried to show that ##f_n## converges uniformly on ##[0,2\pi]##, which is true since ##f_n \rightarrow 0## for ##n \rightarrow \infty## and ##\sigma_n=\mathrm{sup}\left| \frac{\sin\left(\frac{n^2}{n+\frac 15}x\right)}{n^{x^2-3x+3}} \right| \leq \frac{1}{|n^{x^2-3x+3}|} \leq \frac{1}{n^{\frac 34}}\rightarrow 0##. I can't use neither Leibnitz's test nor Abel's test. For Dirichlet's test I would need to show, that ##\sin\left(\frac{n^2}{n+\frac 15}x \right)## has partialy bounded sums...