Integration of rational functions (Section 8.4) Rational function: ratio (quotient) of two polynomials f (x) = Pm (x) , Qn (x) Pm (x) = bm xm + bm−1 xm−1 + ... + b1 x + b0 Qn (x) = an xn + an−1 xn−1 + ... + a1 x + a0 If m < n — proper rational function. We need to find Z Pm (x) dx Qn (x) Idea — represent a proper rational function as a sum of simple terms, ”partial fractions”, for example µ ¶ 1 1 1 1 1 = = + . 1 − x2 (1 − x) (1 + x) 2 1−x 1+x For each term the integral can be evaluated by known methods. Step I. If m ≥ n, transform f into polynomial + proper rational function. Rule #1: Never integrate improper rational function. Transform. Either divide Pm by Qn and find the remaider Pn−1 : Pm (x) = g(x)Qn (x) + Pn−1 (x) or sequentially eliminate from P all terms with powers from m to n. Example: improper rational function f (x) = P4 (x) = x4 − x2 , x4 x2 −x −1 x4 x4 − x2 x2 + x + 1 Q2 (x) = x2 + x + 1 −x2 +x3 +x2 −x3 −2x2 −x3 −x2 −x −x2 +x −x2 −x −1 2x +1 MATH115 T2 Winter 04, A.Potapov (= P4 (x) − x2 Q2 (x)) (= P4 (x) − (x2 Q2 (x) − xQ2 (x))) (= P4 (x) − (x2 Q2 (x) − xQ2 (x) − 1Q2 (x))) — 1 — www.math.ualberta.ca/∼apotapov/math115.htm Therefore, we can conclude that ³ ´ P4 (x) = x2 − x − 1 Q2 + 2x + 1. Check it: ³ x2 − x − 1 ´³ ´ x2 + x + 1 + 2x − 1 = x4 − (x + 1)2 + 2x + 1 = = x4 − x2 − 2x − 1 + 2x + 1 = x4 − x2 . Now ³ ´ f (x) = x2 − x − 1 + 2x + 1 . +x+1 x2 Step 2. Represent Qn (x) as a product of standard factors (ax − b)r and ³ ´r ax2 + bx + c Example: quadratic equation Q2 = ax2 + bx + c. a) b2 − 4ac > 0, Q2 = a (x − α1 ) (x − α2 ) = (ax − aα1 ) (x − α2 ). α1 and α2 are the roots of the quadratic equation ax2 + bx + c = 0, α1 = −b + √ b2 − 4ac , 2a b) b2 − 4ac = 0, Q2 = a (x − α1 )2 ³ c) b2 − 4ac < 0, Q2 = a x2 + ab x + c a α2 = −b − √ b2 − 4ac 2a ´ General case (supplementary) (1) if x can take complex values z = α + iβ, i2 = −1, then there is a theorem: each polynomial Qn (x) of degree n always has n roots (some of the roots may coincide). The polynomial can be rewriten in the form Qn (x) = an xn + an−1 xn−1 + ... + a1 x + a0 = an (x − z1 ) (x − z2 ) ... (x − zn ) So, every simple root contributes a factor (x − zk ), each repeated r times root — (x − zk )r (2) All coefficients ak are real numbers, then complex roots always appear in pairs: if α + iβ is a root, then α − iβ is a root too. Two such roots contribute a factor (x − α − iβ)(x − α + iβ) = (x − α)2 − (iβ)2 = x2 − 2αx + α2 + β 2 This is a quadratic polynomial with the discriminant 4α2 − 4 (α2 + β 2 ) = −4β 2 < 0. (3) if the root α + iβ is repeated r times, then α − iβ is repeated r times too. Then Qn (x) has a factor MATH115 T2 Winter 04, A.Potapov ³ x2 − 2αx + α2 + β 2 — 2 — ´r www.math.ualberta.ca/∼apotapov/math115.htm Conclusion: ³ Qn (x) = (a1 x − b1 )r1 (a2 x − b2 )r2 ... (ak x − bk )rk ak+1 x2 + bk+1 x + ck+1 ´rk+1 ³ ... at x2 + bt x + ct ´rt , where b1 /a1 , b2 /a2 , ..., bk /ak — real roots of the equation Q(x) = 0, r1 + r2 + ... + rk + 2rk+1 + 2rk+2 + ... + 2rr = n. Step 3. express proper fraction P/Q as a sum of partial fractions Pm (x) = (a1 x − b1 ) (a2 x − b2 ) ... (ak x − bk ) (ak+1 x2 + bk+1 x + ck+1 )rk+1 ... (at x2 + bt x + ct )rt r1 r2 rk = X (Partial Fractions) Each factor in Qn (x) generates ri terms in the sum of partial fractions: Factor Terms comment A (ax − b) 1 term ax − b A2 A2 A1 + r terms (ax − b)r 2 + ... + ax − b (ax − b) (ax − b)r Ax + B (ax2 + bx + c) 1 term ax2 + bx + c A1 x + B1 A r x + Br r (ax2 + bx + c) + ... + r terms 2 ax + bx + c (ax2 + bx + c)r Unknown coefficients A, B, C to be determined at the next step Example: f (x) = 1 = 2 à x7 − 3x4 = 2 (x + 1) (x − 1)3 (x2 + x + 1)2 A D Ex + F B C Gx + H + + + 2 2 + 3 + 2 x + 1 x − 1 (x − 1) x + x + 1 (x + x + 1)2 (x − 1) ! where A...H are unknown coefficients. Note: the number of unknown coefficients equals to the power of the polynomial in denominator — here 8. Step 4. Find the coefficients for each partial fraction • Bring the sum of partial fractions to common denominator • Find the polynomial in the numerator, • Coefficients at each power of x of this polynomial should be equal to that of Pm (x). This gives the system of equations for the coefficients at partial fractions MATH115 T2 Winter 04, A.Potapov — 3 — www.math.ualberta.ca/∼apotapov/math115.htm • Solve the system and find the coefficients. Example. 2x2 + 1 A Bx + C = + = (x − 1) (x2 + x + 1) x − 1 x2 + x + 1 = A (x2 + x + 1) + (Bx + C) (x − 1) = (x − 1) (x2 + x + 1) Ax2 + Ax + A + Bx2 + Cx − Bx − C = = (x − 1) (x2 + x + 1) (A + B) x2 + (A − B + C) x + A − C = . (x − 1) (x2 + x + 1) We need that for any x (A + B) x2 + (A − B + C) x + A − C = 2x2 + 1 therefore coefficients at the same powers of x must coincide: (a) A + B = 2, (b) A − B + C = 0, (c) A − C = 1. This is a system of equations for A, B, C. Standard way of solving: From equation (a) A = 2 − B, substitute this into two other equations: (b) 2 − B − B + C = 2 − 2B + C = 0, (c) 2 − B − C = 1. or (b) 2B − C = 2, (c) B + C = 1. From (c) B = 1 − C, substitute into (b): (b) 2 − 2C − C = 2, 3C = 0, C = 0. Knowing C we can find B = 1 − C = 1, and A = 2 − B = 1. There may be faster ways. For example, add all 3 equations, then A + B + A − B + C + A − C = 3, 3A = 3, A = 1, then from (a) B = 2 − A = 1, (c) C = A − 1 = 0. Finally, 1 x 2x2 + 1 = + 2 2 (x − 1) (x + x + 1) x−1 x +x+1 MATH115 T2 Winter 04, A.Potapov — 4 — www.math.ualberta.ca/∼apotapov/math115.htm The cross-out method Some coefficients, corresponding to the highest degree of (ax − b)r can be determined without solving a system of equations. Example. Partial fractions give the relation with 3 unknown constants 2x2 + 1 A Bx + C = + 2 . 2 (x − 1) (x + x + 1) x−1 x +x+1 Multiply this equality by (x − 1) 2x2 + 1 Bx + C = A + (x − 1) 2 . 2 (x + x + 1) x +x+1 There is no singularity at x = 1 any longer. Let us set x = 1, 2+1 = 1 = A. (1 + 1 + 1) At x = 1 terms containing B and C vanished, and we obtain the value of A. Unfortunately, B and C cannot be determined this way — you need complex numbers to do this. Example. Let us consider a general case, Pm (x)/Qn (x). Suppose that Qn (x) = (ax − b)r Sn−r (x), where Sn−r (x) does not contain factors (ax − b). Then Pm (x) Pm (x) = = Qn (x) (ax − b)r Sn−r (x) = Ar A1 Ar−1 + (other terms) . r + r−1 + . . . + (ax − b) (ax − b) (ax − b) Multiply by (ax − b)r , Pm (x) = Ar + Ar−1 (ax − b) + . . . + A1 (ax − b)r−1 + (ax − b)r (other terms) . Sn−r (x) Now set x = b/a, all (ax − b) = 0, what remains is Ar = Pm ( ab ) . Sn−r ( ab ) Other coefficients Ar−1 to A1 cannot be determined this way. Coefficients corresponding to quadratic r terms (ax2 + bx + c) cannot be determined either. This methor can be briefly formulated as follows: to determine Ar 1) ”cross out” the term (ax − b)r from the denominator in P (x)/Q(x) 2) set x = b/a. MATH115 T2 Winter 04, A.Potapov — 5 — www.math.ualberta.ca/∼apotapov/math115.htm Step 5. Integrating For each of the partial fractions evaluate the integral and obtain the final answer. For each partial fraction the integral can be evaluated by known methods: Z dx 1 Z d (ax − b) 1 = = ln |ax − b| + C, ax − b a ax − b a (a) Z (b) dx 1Z 1 = (ax − b)−r d (ax − b) = − (ax − b)−(r−1) + C, r (ax − b) a a (r − 1) r > 1, Z Ax + B dx = ax2 + bx + c here denominator has to be transformed to a standard form without linear term, (c) u=x+ b , 2a ax2 + bx + c = au2 + aβ 2 , ³ 1 Z Au + B − = a u2 + β 2 Ab 2a β2 = 4ac − b2 > 0, 4a2 ´ du = à ! AZ u 1 Ab Z 1 = du + B− du. 2 2 2 a u +β a 2a u + β2 Evaluate the integrals separately Z h i 1Z 1 u 2 2 du = v = u + β = dv = u2 + β 2 2 v ! à ´ 1 ³ 1 ax2 + bx + c = ln u2 + β 2 + C = ln + C, 2 2 a Z 1 β Z 1 du = [u = βv] = dv = 2 2 2 u +β β 1 + v2 à 2ax + b 1 1 = tan−1 v + C = tan−1 √ β β 4ac − b2 ! +C The last one we consider only schematically: Z (d) Ax + B 1 Z A1 u + B1 dx = r du = (ax2 + bx + c)r a (u2 + β 2 )r A1 Z u B1 Z 1 = r du r du + r 2 2 2 a (u + β ) a (u + β 2 )r In the first integral we use substitution v = u2 + β 2 , in the second u = β tan θ, so Z Z Z 1 1−2r 2r 2 1−2r cos θ sec θdθ = β cos2r−2 θ dθ. r du = β 2 2 (u + β ) This integral can be evaluated by successive application of double angle identity. MATH115 T2 Winter 04, A.Potapov — 6 — www.math.ualberta.ca/∼apotapov/math115.htm Examples 1. Z 1 dx 1 − x2 1 1 A B = = + 2 1−x (1 − x) (1 + x) 1−x 1+x Here both coefficients can be obtained by cross-out method: 1 A= 1+x ¸ x=1 1 = , 2 1 B= 1−x ¸ = x=−1 1 2 Z Z 1 1Z 1 1 dx = dx + dx = 2 1−x 2 1−x 1+x ¯ ¯ 1 1 ¯¯ 1 + x ¯¯ 1 + C. = − ln |1 − x| + ln |1 + x| + C = ln ¯ 2 2 2 1 − x¯ 2. Z Z Z dθ dx dx = [x = sin θ] = 3 = 5 2 cos θ (1 − x ) (1 − x)3 (1 + x)3 1 A B C E F D + 3 = 3 + 2 + 3 + 2 + 1 − x (1 + x) 1+x (1 − x) (1 + x) (1 − x) (1 − x) (1 + x) 3 A and D can be determined by cross-out: 1 A= (1 + x)3 # x=1 1 = , 8 1 B= (1 − x)3 # x=−1 1 = . 8 For other coefficients it is necessary to solve a system of equations. 1 (1 + x)3 + B (1 − x) (1 + x)3 + C (1 − x)2 (1 + x)3 + 8 1 + (1 − x)3 + E (1 + x) (1 − x)3 + F (1 + x)2 (1 − x)3 = 1 8 ´ ³ ´³ ´ ³ ´2 1³ 1 + 3x2 + B 1 − x2 1 + x2 + 2x + C 1 − x2 (1 + x) + 4 ³ +E 1 − x2 ´³ ´ ³ 1 + x2 − 2x + F 1 − x2 ³ ´ ´2 (1 − x) = 1 ³ ´ B 1 − x4 + 2x − 2x3 + C 1 − 2x2 + x4 + x − 2x3 + x5 + ³ ´ ³ ´ +E 1 − x4 − 2x + 2x3 + F 1 − 2x2 + x4 − x + 2x3 − x5 = 3 3 2 − x 4 4 3 4 1 x : 2B + C − 2E − F = 0 3 x2 : −2C − 2F = − 4 3 x : −2B − 2C + 2E + 2F = 0 x0 : MATH115 T2 Winter 04, A.Potapov B+C +E+F = — 7 — www.math.ualberta.ca/∼apotapov/math115.htm x4 : −B + C − E + F = 0 x5 : So C = F , C −F =0 3 3 (2) 2C + 2F = 4C = , C=F = , 4 16 (3) B + C = E + F, B=E 3 3 3 3 3 (0) B + E = 2B = − C − F = − = , B=E= . 4 4 8 8 16 Z dx 1Z dx 3 Z dx 3 Z dx = + + + 8 (1 − x)3 16 (1 − x)2 16 1 − x (1 − x2 )3 1Z dx 3 Z dx 3 Z dx + + + = 8 (1 + x)3 16 (1 + x)2 16 1 + x 3 1 1 1 3 = − ln |1 − x| − 2 + 16 (1 − x) 16 1 − x 16 1 3 1 3 1 − + ln |1 + x| + C = 2 − 16 (1 + x) 16 1 + x 16 ¯ ¯ 3 1 + x − (1 − x) 1 (1 + x)2 − (1 − x)2 3 ¯¯ 1 + x ¯¯ + = + ln +C = 16 16 1 − x2 16 ¯ 1 − x ¯ (1 − x2 )2 ¯ ¯ 3x 3 ¯¯ 1 + x ¯¯ x + + ln ¯ = + C = ... (x = sin θ) 1 − x¯ 4 (1 − x2 )2 8 (1 − x2 ) 16 3. Z 3x2 + 2 dx (x − 1) (x2 + 2x + 2) 3x2 + 2 A Bx + C = + 2 , 2 (x − 1) (x + 2x + 2) x − 1 x + 2x + 2 # 3x2 + 2 A= 2 = 1, (x + 2x + 2) x=1 Bring to common denominator, J= 3x2 + 2 = x2 + 2x + 2 + (Bx + C) (x − 1) , Bx2 + Cx − Bx − C = 2x2 − 2x, B = 2, C − B = −2, −C = 0. Z Z 2 3x + 2 1 x dx = dx + 2 dx 2 2 (x − 1) (x + 2x + 2) x−1 x + 2x + 2 Z 1 dx = ln |x − 1| + C, x−1 Z Z x x+1−1 dx = dx = [u = x + 1] 2 x + 2x + 2 (x + 1)2 + 1 Z Z Z ´ 1 ³ 2 u−1 udu du = du = − = ln u + 1 − tan−1 u + C. 2 2 2 u +1 u +1 u +1 2 ³ ´ 2 J = ln |x − 1| + ln x + 2x + 2 − 2 tan−1 (x + 1) + C Z MATH115 T2 Winter 04, A.Potapov — 8 — www.math.ualberta.ca/∼apotapov/math115.htm
© Copyright 2026 Paperzz