Laplace transforms are one of the ways of solving linear ordinary differential equations (Linear ODEs) with constant coefficients. This technique allows us to transform a Linear ODE into a linear algebraic equation.
Definition
The unilateral Laplace transform is defined by

Given the integral converges. A necessary condition for this integral to converge is

Examples
Example 1
Consider the following initial value problem

where
is constant, subject to

To solve this problem using laplace transform, first apply laplace transform to both sides of the equation, obtaining:

Or

The integral on the right hand side is


![{\displaystyle =\lim _{b\to \infty }\left[{\frac {1}{a-s}}e^{(a-s)t}-{\frac {1}{a-s}}e^{(a-s)(0)}\right]}](https://wikimedia.org/api/rest_v1/media/math/render/svg/b822a7ac3039c3ff22c41fc4c1e1347f2f3df89f)
![{\displaystyle =\lim _{b\to \infty }\left[{\frac {1}{a-s}}e^{(a-s)t}-{\frac {1}{a-s}}\right]}](https://wikimedia.org/api/rest_v1/media/math/render/svg/93a04e06bac899edc365fa01b024aef47df4e116)
If
,
![{\displaystyle \lim _{b\to \infty }\left[{\frac {1}{a-s}}e^{(a-s)t}-{\frac {1}{a-s}}\right]={\frac {1}{a-s}}}](https://wikimedia.org/api/rest_v1/media/math/render/svg/2dd9fac1934bbf506a5b7ddf89053ad7140d888a)
For the left side, if we apply integration by parts,

Substitution into the left side will get



For the transformation to converge,

Therefore, substituting the initial condition y(0)=0 the left side becomes

Equating the two sides of the equation:


If
, we can use partial fractions to changet the right side into

and the solution is obtained by noticing:

Substituting the inverse transform we have

![{\displaystyle \int _{0}^{\infty }e^{-st}y\,dt=\int _{0}^{\infty }e^{-st}\left[{\frac {1}{a+1}}\left(e^{at}-e^{-t}\right)\right]\,dt}](https://wikimedia.org/api/rest_v1/media/math/render/svg/9243917a511e31a10f0f8650b991182763c6b529)
Which leads to the answer

References
- D. Lomen and D. Lovelock, Differential Equations Graphics. Model. Data., John Wiley and Sons, Toronto, 1999.
- Laplace transform on Wolfram Mathworld