It seems to be a hard integral but it is not

แชร์
ฝัง
  • เผยแพร่เมื่อ 18 มิ.ย. 2024
  • Mis-1319
    #calculus #improperintegrals #substitution #properties #algebraic #manipulation #cipher

ความคิดเห็น • 2

  • @slavinojunepri7648
    @slavinojunepri7648 6 วันที่ผ่านมา

    Gorgeous!
    This integral would very difficult without the inverse function substitution.

  • @julianwang7987
    @julianwang7987 7 วันที่ผ่านมา

    Although the solution given in the video is by far the simplest, it may be of interest for some to see how Feynman's trick can apply to this question.
    Rewrite the original integral as integral_0^inf{dx (ln(1+x^3)-ln(1+x))/(1+x^2)/ln(x)} = F(3)-F(1), where F(a) = integral_0^inf{dx ln(1+x^a)/(1+x^2)/ln(x))}. Therefore, F'(a) = integral_0^inf{dx x^a/(1+x^2)/(1+x^a)} = integral_0^(pi/2){ds tan^a(s)/(1+tan^a(s))} = integral_0^(pi/2){ds sin^a(s)/(sin^a(s)+cos^a(s))} = integral_0^(pi/2){ds cos^a(s)/(sin^a(s)+cos^a(s))} = 1/2*integral_0^(pi/2){ds (sin^a(s)+cos^a(s))/(sin^a(s)+cos^a(s))} = pi/4. Thus, I = F(3)-F(1) = integral_1^3{da pi/4} = pi/2