Unfortunately, standard integration techniques will not help you solve this integral. The actual anti-derivative of this function is huge (according to Wolfram alpha at least, see: https://www.wolframalpha.com/input/?i=integral+arctan%281%2F%28cosh%28x%29%29%29 ). To combat this, will we use a method called Feynman Integration (named after the Physicist Richard Feynman. Although the actual rule was discovered by Leibniz - who independently discovered Calculus).
Let
$${I(t)=\int_{0}^{\infty}\arctan\left(\frac{t}{\cosh(x)}\right)dx}$$
So we have defined a function in terms of our integral. Using the Leibniz rule for integration we get
$${I'(t)=\int_{0}^{\infty}\frac{\text{sech}(x)}{t^2\text{sech}^2(x) + 1}dx}$$
(to take the derivative, you take take the partial derivative of the inside :D). The inner function now has an elementary anti-derivative; namely
$${\int\frac{\text{sech}(x)}{1+t^2\text{sech}^2(x)}dx=\frac{-\arctan\left(\sqrt{t^2 + 1}\text{csch}(x)\right)}{\sqrt{1+t^2}} + C}$$
Hence the integral for ${I'(t)}$ can be found by taking limits:
$${\int_{0}^{\infty}\frac{\text{sech}(x)}{t^2\text{sech}^2(x) + 1}dx=\lim_{x\rightarrow \infty}\frac{-\arctan\left(\sqrt{t^2 + 1}\text{csch}(x)\right)}{\sqrt{1+t^2}} - \lim_{x\rightarrow 0}\frac{-\arctan\left(\sqrt{t^2 + 1}\text{csch}(x)\right)}{\sqrt{1+t^2}}}$$
$${\Rightarrow I'(t) = \frac{\pi}{2}\frac{1}{\sqrt{1+t^2}}}$$
So to find ${I(t)}$ we now simply integrate with respect to ${t}$ and find the constant. This gives us
$${I(t)=\frac{\pi}{2}\int\frac{1}{\sqrt{t^2 + 1}}dt=\frac{\pi}{2}\sinh^{-1}(t) + C}$$
(${\int\frac{1}{\sqrt{1+t^2}}dt}$ is just a known integral).
But ${I(0)=0\Rightarrow C=0}$ (since ${\sinh^{-1}(0)=0}$), hence
$${\int_{0}^{\infty}\arctan\left(\frac{n}{\cosh(x)}\right)dx=\frac{\pi}{2}\sinh^{-1}(n)}$$, but
$${\sinh^{-1}(n)=\ln\left(\sqrt{n^2 + 1} + n\right)}$$
and so indeed
$${\int_{0}^{\infty}\arctan\left(\frac{n}{\cosh(x)}\right)dx=\frac{\pi}{2}\ln\left(\sqrt{n^2 + 1} + n\right)}$$