Transcript a n (x)

191
4-4 Undetermined Coefficients –
Superposition Approach
This section introduces some method of “guessing” the particular
solution.
4-4-1 方法適用條件
(1)
(2)
Suitable for linear and constant coefficient DE.
an y ( n )  x   an1 y ( n1)  x  
 a1 y( x )  a0 y  x   g  x 
(3) g(x), g'(x), g'' (x), g'''(x), g(4)(x), g(5)(x), ………contain
finite number of terms.
192
4-4-2 方法
把握一個原則:
g(x) 長什麼樣子,particular solution 就應該是什麼樣子.
記熟下一頁的規則
(計算時要把 A, B, C, … 這些 unknowns 解出來)
193
Trial Particular Solutions
(from text page 143)
g(x)
Form of yp
1 (any constant)
A
5x + 7
Ax + B
3x2 – 2
Ax2 + Bx + C
x3 – x + 1
Ax3 + Bx2 + Cx + E
sin4x
Acos4x + Bsin4x
cos4x
Acos4x + Bsin4x
e5x
Ae5x
(9x – 2)e5x
(Ax + B)e5x
x2e5x
(Ax2 + Bx + C)e5x
e3xsin4x
Ae3xcos4x + Be3xsin4x
5x2sin4x
(Ax2 + Bx + C)cos4x + (Ex2 + Fx + G)sin4x
xe3xcos4x
(Ax + B)e3xcos4x + (Cx + E)e3xsin4x
It comes from the “form rule”. See page 198.
194
g  x   e 2 x  xe3 x
yp = ?
g  x   cos  x   x 2 sin  2 x 
yp = ?
g  x   cosh  2 x 
yp = ?
195
4-4-3 Examples
Example 2
y  y  y  2sin 3x
(text page 141)
Step 1: find the solution of the associated homogeneous equation
Guess
Step 2: particular solution
y p  A cos3x  B sin 3x
yp  3 A sin 3x  3B cos3x
yp  9 A cos3x  9 B sin 3x
yp  yp  y p  (8 A  3B)cos3x  (3 A  8B)sin 3x  2sin 3x
8 A  3B  0

 3 A  8B  2
Step 3: General solution:
A = 6/73, B = 16/73
y p  6 cos3 x  16 sin 3 x
73
73


y  e x / 2  c1 cos 3 x  c2 sin 3 x   6 cos3x  16 sin 3x
2
2  73
73

Example 3
y  2 y  3 y  4 x  5  6 xe
2x
(text page 142)
196
Step 1: Find the solution of
y  2 y  3 y  0.
yc  c1e3 x  c2e  x
Step 2: Particular solution
y  2 y  3 y  4 x  5
guess
y p1  Ax  B
yp1  A
yp1  0
3 Ax  2 A  3B  4 x  5
A   4 , B  23
3
9
y p1   4 x  23
3
9
y  2 y  3 y  6 xe2 x
guess
y p2  Cxe2 x  Ee2 x
yp2  2Cxe 2 x  Ce 2 x  2 Ee 2 x
yp2  4Cxe 2 x  4Ce 2 x  4 Ee 2 x
3Cxe 2 x   2C  3E  e 2 x  6 xe 2 x
C  2, E   4
3
y p2  (2 x  4 )e 2 x
3
Particular solution
y p  y p1  y p2   4 x  23  (2 x  4 )e x
3
9
3
Step 3: General solution
y  yc  y p
y  c1e3 x  c2e  x  4 x  23  (2 x  4 )e 2 x
3
9
3
197
4-4-4 方法的解釋
Form Rule: yp should be a linear combination of g(x), g'(x),
g'' (x), g'''(x), g(4)(x), g(5)(x), …………….
Why? 如此一來,在比較係數時才不會出現多餘的項
198
When g(x) =
199
xn
x n  x n1  x n2  x n3 
y p  An x n  An1 x n1  An2 x n2 
When g(x) = cos kx
cos kx  sin kx
y p  A1 cos kx  A2 sin kx
When g(x) = exp(kx)
e kx
y p  A exp(kx)
1 0
 A0
When g(x) =
200
xnexp(kx)
g   x   nx n1e kx  kx n e kx
g   x   n(n  1) x n2e kx  2nkx n1e kx  k 2 x ne kx
g   x   n(n  1)(n  2) x n3e kx  3kn(n  1) x n2e kx
3k 2 nx n1e kx  k 3 x ne kx
:
:
會發現 g(x) 不管多少次微分,永遠只出現
x nekx , x n1ekx , x n2ekx , x n3ekx ,
y p  cn x nekx  cn1 x n1ekx  cn2 x n2e kx 
, ekx
 c0e kx
201
4-4-5 Glitch of the method:
Example 4
y  5 y  4 y  8e x
(text page 142)
Particular solution guessed by Form Rule:
y p  Ae x
yp  5 yp  4 y p  Ae x  5 Ae x  4 Ae x  8e x
0  8e x
Why?
(no solution)
202
Glitch condition 1: The particular solution we guess belongs to the
complementary function.
For Example 4
y  5 y  4 y  8e x
x
4x
Complementary function yc  c1e  c2e
Ae x  yc
解決方法:再乘一個 x
y p  Axe
x
yp  Axe x  Ae x
yp  Axe x  2 Ae x
yp  5 yp  4 y p  3 Ae x  8e x
y p   8 xe x
3
y  c1e x  c2e 4 x  8 xe x
3
A  8/ 3
Example 7
y  2 y  y  e x
203
(text page 144)
yc  c1e x  c2 xe x
From Form Rule, the particular solution is Aex
Ae x  yc
Axe x  yc
y p  Ax 2e x
如果乘一個 x 不夠,則再乘一個 x
yp  ( Ax 2  2 Ax)e x
yp  ( Ax 2  4 Ax  2 A)e x
yp  2 yp  y p  2 Ae x  e x
y p  x 2e x / 2
y  c1e x  c2 xe x  x 2e x / 2
A  1/ 2
Example 8 (text page 145)
y  y  4 x  10sin x
204
y    0
Step 1
yc  c1 cos x  c2 sin x
Step 2
y p  Ax  B  Cx sin x  Ex cos x
y     2
注意: sinx, cosx 都要
乘上 x
y p  4 x  5 x cos x
Step 3
y  c1 cos x  c2 sin x  4 x  5 x cos x
Step 4
Solving c1 and c2 by initial conditions
(最後才解 IVP)
c1  9
y    c1  4  5  0
y  c1 sin x  c2 cos x  4  5cos x  5 x sin x
c2  7
y    c  9  2
2
y  9 cos x  7sin x  4 x  5 x cos x
205
Example 11 (text page 146)
y (4)  y  1  x 2e x
yc  c1  c2 x  c3 x 2  c4e  x
From Form Rule
y p  A  Bx 2e x  Cxe x  Ee x
修正
yp 只要有一部分和 yc 相
同就作修正
y p  Ax3  Bx3e x  Cx 2e x  Exe  x
乘上 x3
2 x
x
x
If we choose y p  A  Bx e  Cxe  Ee
乘上 x
x
x
2 x

y (4)

y


2
Bxe

(6
B

C
)
e

1

x
e
p
( p)
沒有 1, x2ex 兩項,不能比較係數,無解
206
If we choose y p  Ax3  Bx 2e x  Cxe x  Ee x
x
x
2 x

y (4)

y

6
A

2
Bxe

(6
B

C
)
e

1

x
e
p
( p)
沒有 x2ex 這一項,不能比較係數,無解
If we choose y p  Ax3  Bx3e x  Cx 2e x  Exe  x

y (4)
p  y( p )
 6 A  3Bx 2e  x  18 B  2C  xe  x  (18 B  6C  E )e  x
 1  x 2e x
A = 1/6, B = 1/3, C = 3, E = 12
y p  1 x3  1 x3e  x  3x 2e  x  12 xe  x
6
3
y  c1  c2 x  c3 x 2  c4e  x  1 x 3  1 x 3e  x  3x 2e  x  12 xe  x
6
3
207
Glitch condition 2: g(x), g'(x), g'' (x), g'''(x), g(4)(x), g(5)(x), ……………
contain infinite number of terms.
If g(x) = ln x
ln x 
1
1
1
 2 3
x
x
x
If g(x) = exp(x2)
g   x   2 xe x
2
g   x   (4 x  2)e
3
x2

g  x   (8 x  12 x)e
2
:
:
x2
208
4-4-6 本節需要注意的地方
(1) 記住 Table 4.1 的 particular solution 的假設方法
(其實和 “form rule” 有相密切的關聯)
(2) 注意 “glitch condition”
另外,“同一類” 的 term 要乘上相同的東西 (參考 Example
11)
(3) 所以要先算 complementary function,再算 particular solution
(4) 同樣的方法,也可以用在 1st order 的情形
(5) 本方法只適用於 linear, constant coefficient DE
209
4-5 Undetermined Coefficients –
Annihilator Approach
For a linear DE:
an y ( n )  x   an1 y ( n1)  x  
 a1 y( x )  a0 y  g  x 
Annihilator Operator:
能夠「殲滅」 g(x) 的 operator
4-5-1 方法適用條件
(1) Linear, (2) Constant coefficients
(3) g(x), g'(x), g'' (x), g'''(x), g(4)(x), g(5)(x), ………contain
finite number of terms.
210
4-5-2 Find the Annihilator
Example 1: (text page 150)
g  x   1  5 x 2  8 x3
annihilator: D4
g  x   e 3x
annihilator: D + 3
k
d
Dk g  x   k g  x 
dx
d g  x   3g  x   0
dx
g  x   4e 2 x  10 xe 2 x
211
annihilator: (D − 2)2
(D − 2)2 = D2 − 4D + 4
d 2 g  x  4 d g  x   4g  x   0
dx
dx 2
註:當 coefficient 為 constants 時,function of D 的計算方式
和 function of x 的計算方式相同
(x − 2)2 = x2 − 4x + 4
 (D − 2)2 = D2 − 4D + 4
212
General rule 1:
If
g  x    an x n  an1x n1 
then the annihilator is  D   
 a0  e x
n 1
注意: annihilator 和 a0, a1, …… , an 無關
只和 , n 有關
213
General rule 2:
n
n1
If g  x    an x  an1x 
 a0  e x  b1 cos  x  b2 sin  x 
b1  0 or b2  0
2
2
2
then the annihilator is  D  2 D      
Example 2: (text page 151)
annihilator
Example 5: (text page 154)
annihilator
Example 6: (text page 155)
annihilator
n 1
g  x   5e  x cos 2 x  9e  x sin 2 x
D2  2D  5
g  x   x cos x  cos x
2
 D  1
g  x   10e 2 x cos x
2
D2  4D  5
214
General rule 3:
If g(x) = g1(x) + g2(x) + …… + gk(x)
Lh[gh(x)] = 0 but Lh[gm(x)]  0 if m  h,
then the annihilator of g(x) is the product of Lh (h = 1 ~ k)
Lk Lk 1
Proof:
Lk Lk 1
 Lk Lk 1
L2 L1
L3 L2 L1  g1  g 2  g 3 
L3 L2 L1 g1  Lk Lk 1
 gk 
L3 L2 L1 g 2 
Lk Lk 1
L3 L2 L1 g3 
 Lk Lk 1
Lk Lk 1
L3 L2 L1g1  Lk Lk 1
L3 L2  L1g1   0
Lk Lk 1
L3 L2 L1g2  Lk Lk 1
L3 L1  L2 g2   0
L3 L2 L1 g k
(因為 L1, L2 為 linear DE with constant coefficient,
L1L2 = L2L1 )
215
Similarly,
Lk Lk 1
L4 L3 L2 L1g3  Lk Lk 1
Lk Lk 1
:
:
L4 L3 L2 L1g3  Lk 1
L4 L2 L1  L3 g3   0
L4 L3 L2 L1  Lk g k   0
Therefore,
Lk Lk 1
L3 L2 L1  g1  g 2  g 3 
 000
0
0
 gk 
216
Example 7 (text page 154)
g  x   5 x 2  6 x  4 x 2e 2 x  3e5 x
annihilator: D − 5
annihilator: D3
annihilator: (D − 2)3
annihilator of g(x): D3 (D − 2)3 (D − 5)
217
4-5-3 Using the Annihilator to Find the Particular Solution
Step 2-1 Find the annihilator L1 of g(x)
Step 2-2 如果原來的 linear & constant coefficient DE 是
L y  g  x
那麼將 DE 變成如下的型態:
L1  L  y   L1  g  x   0
(homogeneous linear & constant coefficient DE)
(n)
( n 1)
 x 
註: If an y  x   an1 y
n
n1
then L  an D  an1D 
 a1 y( x )  a0 y  g  x 
 a1D  a0
218
Step 2-3 Use the method in Section 4-3 to find the solution of
L1  L  y   0
Step 2-4 Find the particular solution.
The particular solution yp is a solution of
L1  L  y   0
but not a solution of
L y  0
(Proof): Since L  y p   g ( x) , if g(x)  0, L  y p  should be nonzero.
Moreover, L1  L  y p    L1  g  x   0.
Step 2-5 Solve the unknowns
219
solutions of
L1  L  y   0
particular solution yp
solutions of
L y  0
particular solution yp  solutions of L1  L  y   0
 solutions of L  y   0
本節核心概念
4-5-4 Examples
220
Example 3 (text page 152)
y  3 y  2 y  4 x 2
Step 1: Complementary function
(solution of the associated homogeneous function)
yc  c1e x  c2e2 x
Step 2-1: Annihilation: D3
L1  L  y   L1  g  x   0
Step 2-2:
D3 ( D2  3D  2) y  0
Step 2-3: auxiliary function m3 (m2  3m  2)  0
roots: m1 = m2 = m3 = 0, m4 = −1, m5 = −2
移除和 complementary
Solution for L1  L  y   0 :
function 相同的部分
2
x
2 x
y  d1  d 2 x  d3 x  d 4e  d5e
Step 2-4: particular solution y p  A  Bx  Cx
Step 2-5:
2
yp  B  2Cx
yp  2C
yp  3 yp  2 y  2Cx 2  (2 B  6C ) x  (2 A  3B  2C )  4 x 2
2C  4


 2 B  6C  0
 2 A  3B  2C  0

C2
B  6
A7
y p  7  6x  2x2
x
2 x
2
Step 3: y  yc  y p  c1e  c2e  7  6 x  2 x
221
222
Example 4 (text page 153)
y  3 y  8e3 x  4sin x
Step 1: Complementary function
From auxiliary function, m2 − 3m = 0, roots: 0, 3
yc  c1  c2e3 x
Step 2-1: Find the annihilator
D−3
annihilate 8e3 x
(D2 + 1)
annihilate 4sin x but cannot annihilate 8e3 x
but cannot annihilate 4sin x
(D − 3)(D2 + 1) is the annihilator of 8e3 x  4sin x
Step 2-2:
 D  3 ( D2  1)  D2  3D  y  0
2
2
Step 2-3: auxiliary function:  m  3 (m  1)  m  3m 
 m  m  3 (m 2  1)  0
223
易犯錯的地方
2
solution of
 D  3 ( D2  1)  D2  3D  y  0 :
y  d1  d 2e3 x  d3 xe3 x  d 4 cos x  d5 sin x
Step 2-4: particular solution
y p  d3 xe3 x  d 4 cos x  d5 sin x
代回原式
並比較係數
8 xe3 x  6 cos x  2 sin x
y

Step 2-5:
p
3
5
5
Step 3: general solution y  c1  c2e3 x  8 xe3 x  6 cos x  2 sin x
3
5
5
4-5-5 本節要注意的地方
(1) 所以要先算 complementary function,再算 particular solution
(2) 若有兩個以上的 annihilator,選其中較簡單的即可
(3) 計算 auxiliary function 時有時容易犯錯
(4) L1  L  y   0 的解和 L  y   0 的解不一樣。
(5) 這方法,只適用於 constant coefficient linear DE
(因為,還需借助 auxiliary function)
224
225
The thing that can be done by the annihilator approach can always
be done by the “guessing” method in Section 4-4, too.
226
4-6 Variation of Parameters
4-6-1 方法的限制
The method can solve the particular solution for any linear DE
(1) May not have constant coefficients
(2) g(x) may not be of the special forms
an  x  y ( n )  x   an1  x  y ( n1)  x  
 a1  x  y( x )  a0  x  y  g  x 
227
4-6-2 Case of the 2nd order linear DE
a2  x  y( x)  a1  x  y( x)  a0  x  y  g  x 
associated homogeneous equation: an  x  y( x)  a1  x  y( x)  a0  x  y  0
Suppose that the solution of the associated homogeneous equation is
c1 y1 ( x)  c2 y2 ( x)
Then the particular solution is assumed as:
y p  u1 ( x) y1 ( x)  u2 ( x) y2 ( x)
(方法的基本精神)
y p  u1 ( x) y1 ( x)  u2 ( x) y2 ( x)
代入原式後,總是可以簡化
228
yp  u1 y1  u1 y1  u2 y2  u2 y2
yp  u1y1  2u1 y1  u1 y1  u2 y2  2u2 y2  u2 y2
代入 y( x)  P  x  y( x)  Q  x  y  f  x 
a0 ( x)
a1 ( x)
P  x 
, Q  x 
,
a2 ( x)
a2 ( x)
zero
g ( x)
f  x 
a2 ( x )
zero
yp  P  x  yp  Q  x  y p  u1  y1  Py1  Qy1   u2  y2  Py2  Qy2   y1u1
2u1 y1  y2u2  2u2 y2  P  y1u1  y2u2 
yp  P  x  yp  Q  x  y p  f  x  ,
y p  u1 y1  u2 y2
簡化
d  y u  y u   P  y u  y u   yu  y u  f  x 
2 2
1 1
2 2
1 1
2 2
dx 1 1
進一步簡化:
假設 y1u1  y2u2  0
y1u1  y2 u2  f  x 
聯立方程式
 y1u1  y2u2  0
    
 y1u1  y2u2  f  x 
229
 y1u1  y2u2  0
    
 y1u1  y2u2  f  x 
y1
where W 
y1
y2
y2
y2 f  x 
W1

u1 

W
W
u2 
W1 
W2 y1 f  x 

W
W
0
f ( x)
y2
y2
y1
W2 
y1
| |: determinant
y p  x   u1  x  y1  x   u2  x  y2  x 
可以和 1st order case (page 58) 相比較
u1   u1  x  dx
u2   u2  x  dx
0
f ( x)
230
4-6-3 Process for the
2nd
231
Order Case
Step 2-1 變成 standard form
y( x)  P  x  y( x)  Q  x  y  f  x 
Step 2-2
y1
W
y1
y2
y2
W1
W
W1 
u2 
0
f ( x)
y2
y2
W2
W
Step 2-3
u1 
Step 2-4
u1   u1  x  dx
Step 2-5
y p  x   u1  x  y1  x   u2  x  y2  x 
u2   u2  x  dx
y1
W2 
y1
0
f ( x)
232
4-6-4 Examples
Example 1 (text page 159)
y  4 y  4 y  ( x  1)e2 x
Step 1: solution of y  4 y  4 y  0 :
yc  c1e2 x  c2 xe2 x
Step 2-2: y p  u1 y1  u2 y2 ,
W
W2 
e2 x
2e
2x
xe2 x
2 xe  e
2x
2x
e2 x
0
2e2 x
( x  1)e2 x
Step 2-3: u1 
y1  e2 x ,
e
4x
W1 
y2  xe2 x
0
xe2 x
( x  1)e2 x
2 xe2 x  e2 x
 ( x  1)e4 x
W1
  x2  x
W
u2 
W2
 x 1
W
 ( x  1) xe4 x
Step 2-4: u1   u1dx   ( x  x)dx   1 x 3  1 x 2  c1
3
2
u2   u2 dx   ( x  1)dx  1 x 2  x  c2
2
2
Step 2-5: y p  ( 1 x3  1 x 2 )e 2 x  ( 1 x 2  x) xe 2 x  ( 1 x 3  1 x 2 )e2 x
3
2
2
6
2
Step 3:
y  c1e 2 x  c2 xe 2 x  ( 1 x3  1 x 2 )e 2 x
6
2
233
4 y  36 y  csc3x
Example 2 (text page 159)
Step 1: solution of
4 y  36 y  0 :
Step 2-1: standard form:
Step 2-2:
W
cos3 x
yc  c1 cos3x  c2 sin 3x
y  9 y  csc3x / 4
sin 3 x
3sin 3 x 3cos3 x
cos3 x
234
3
f ( x)  csc3 x / 4
0
sin 3 x
W1  1
 1/ 4
csc3 x 3cos3 x
4
0
1 cos3 x

 sin 3 x 1 csc3 x 4 sin 3 x
4
W1
W2 1 cos3x
1


u



u


Step 2-3: 1
2
12
W
W 12 sin 3x
x
Step 2-4: u1   12
u2  1 ln sin 3 x
36
1 cos3 x
(未完待續)
注意
dx

12 sin 3 x
W2 
算法
x
1
Step 2-5: y p   12 cos3 x  36 sin 3 x ln sin 3 x
Step 3:
235
y  yc  y p  c1 cos3x  c2 sin 3x  x cos3x  1 sin 3x ln sin 3x
12
36
Note: 課本 Interval (0, /6) 改為(0, /3)
Example 3 (text page 160)
y  y  1/ x
f ( x)  1/ x
Note:
ex
 x dx
沒有 analytic 的解
et
所以直接表示成 
dt
x0 t
x
(複習 page 45)
236
4-6-5 Case of the Higher Order Linear DE
an  x  y ( n )  x   an1  x  y ( n1)  x  
 a1  x  y( x )  a0  x  y  g  x 
Solution of the associated homogeneous equation:
yc  c1 y1 ( x)  c2 y2 ( x)  c3 y3 ( x) 
 cn yn ( x)
The particular solution is assumed as:
y p  u1 ( x) y1 ( x)  u2 ( x) y2 ( x)  u3 ( x) y3 ( x) 
uk ( x) 
Wk
W
uk ( x)   uk ( x)dx
 u n ( x ) yn ( x )
237
uk ( x) 
Wk
W
W
y1
y1
y2
y2
y3
y3
yn
yn
y1
y2
y3
yn
y1( n1)
y2( n1)
y3( n1)
yn( n1)
y1
y1
y2
y2
yk 1
yk 1
0
y1( n2)
y2( n2)
y1( n1)
y2( n1)
0
yk 1
yk 1
yn
yn
yk( n12)
0
yk( n1 2)
yn( n2)
yk( n11)
f ( x)
yk( n11)
yn( n1)
Wk 
f  x   g  x  / an  x 
Wk: replace the kth column of W by
 0 
 0 




 0 


 f ( x ) 
238
g  x
f  x 
an  x 
For example, when n = 3,
0
W1 
0
y2
y2
y3
y3
f ( x)
y2
y3
y1
0
W2  y1
0
y1 f ( x)
y3
y3
y3
y1 y2
W3  y1 y2
y1 y2
0
0
f ( x)
239
4-6-6 Process of the Higher Order Case
Step 2-1 變成 standard form
an1  x  ( n1)
(n)
y  x 
y
 x 
an  x 
a1  x 
a0  x 
g  x


y ( x) 
y
an  x 
an  x 
an  x 
Step 2-2 Calculate W, W1, W2, …., Wn (see page 237)
Step 2-3 u1 
W1
W
u2 
Step 2-4 u1   u1  x  dx
Step 2-5
W2
W
……… un 
Wn
W
u2   u2  x  dx ……. un   un  x  dx
y p  x   u1  x  y1  x   u2  x  y2  x  
 un  x  yn  x 
240
4-6-7 本節需注意的地方
(1)養成先解 associated homogeneous equation 的習慣
(2) 記熟幾個重要公式
(3) 這裡 | | 指的是 determinant
(4) 算出 u1(x) 和 u2(x) 後別忘了作積分
特別要小心
(5) f(x) = g(x)/an(x) (和 1st order 的情形一樣,使用 standard form)
(6) 計算 u1'(x) 和 u2'(x) 的積分時,+ c 可忽略
因為 我們的目的是算particular solution yp
yp 是任何一個能滿足原式的解
(7) 這方法解的範圍,不包含 an(x) = 0 的地方
241
4-7 Cauchy-Euler Equation
4-7-1 解法限制條件
an x n y ( n )  x   an1 x n1 y ( n1)  x  
not constant coefficients
 a1 xy( x )  a0 y  g  x 
k
k
but the coefficients of y(k)(x) have the form of ak x
ak is some constant
associated homogeneous
equation
an x n y ( n )  x   an1 x n1 y ( n1)  x  
 a1 xy( x)  a0 y  0
particular solution
242
4-7-2 解法
Associated homogeneous equation of the Cauchy-Euler equation
an x n y ( n )  x   an1 x n1 y ( n1)  x  
 a1xy( x )  a0 y  0
Guess the solution as y(x) = xm , then
 m  n  1 x mn 
an1 x n1m(m  1)(m  2)
 m  n  2  x mn1 
an2 x n2 m(m  1)(m  2)
 m  n  3 x mn 2 
an x n m(m  1)(m  2)
 a1 xmx m1
 a0 x m  0
an m(m  1)(m  2)
 an1m(m  1)(m  2)
 an2 m(m  1)(m  2)
 m  n  1
 m  n  2
 m  n  3
243
auxiliary function
比較: 和 constant coefficient
時有何不同?
 a1m
 a0  0
k
m!
d
規則:把 x
變成
(m  k )!
dx k
k
4-7-3 For the
2nd
244
Order Case
a2 x 2 y  x   a1 xy( x)  a0 y  0
auxiliary function:
a2m  m  1  a1m  a0  0
roots
m1 
a2  a1 
a2 m2   a1  a2  m  a0  0
 a1  a2   4a2a0
a2  a1 
m2 
2a2
2
 a1  a2 2  4a2a0
[Case 1]: m1  m2 and m1, m2 are real
two independent solution of the homogeneous part:
x m1 and x m2
yc  c1 x m1  c2 x m2
2a2
245
[Case 2]: m1 = m2
Use the method of reduction of order
y1  x m1
a1
dx
a2 x
 P ( x ) dx

e 
e
y2  x   y1  x   2
dx  x m1  2 m1 dx
y1  x 
x

a0
a1

y
(
x
)

y  0,
Note 1: 原式
2
a2 x
a2 x
a2  a1
Note 2: 此時 m1  m2 
2a2
y  x  
a1
P  x 
a2 x
246

y2  x   x
  1
a
 1
a2
m1

e

x
x m1  x
a1
dx
a2 x
2 m1

a1
a2
x

dx  x
a1  a2
a2
m1

e
a1
ln x
a2
x
2 m1
dx  x
m1


a1
a2
x
dx
2 m1
x
dx  x m1  x 1dx  x m1 ln x
If y2(x) is a solution of a homogeneous DE
then c y2(x) is also a solution of the homogeneous DE
m
If we constrain that x > 0, then y2  x 1 ln x
yc  c1 x m1  c2 x m1 ln x
247
[Case 3]: m1  m2 and m1, m2 are the form of
m2    j 
m1    j 
two independent solution of the homogeneous part:
x  j and
x  j
yc  C1 x  j  C2 x  j
x  j  (elnx )  j  e(  j ) ln x  e ln x e j ln x
 x  cos(  ln x)  j sin(  ln x) 
  j

x

x
 cos(  ln x)  j sin(  ln x) 
同理
yc  x [(C1  C2 )cos   ln x   (C1  C2 )sin   ln x ]
yc  x [c1 cos   ln x   c2 sin   ln x ]
248
Example 1 (text page 163)
x 2 y  x   2 xy( x)  4 y  0
Example 2 (text page 164)
4 x 2 y  x   8 xy( x)  y  0
249
Example 3 (text page 165)
4 x y  x   17 y  0
2
y 1  1
y 1   1
2
250
4-7-4 For the Higher Order Case
Process:
auxiliary function
roots
n independent solutions
Step 1-1
Step 1-2
solution of the nth order associated
Step 1-3
homogeneous equation
251
(1) 若 auxiliary function 在 m0 的地方只有一個根
xm0
是 associated homogeneous equation 的其中一個解
(2) 若 auxiliary function 在 m0 的地方有 k 個重根
x m0 , x m0 ln x , x m0 (ln x) 2 ,
, x m0 (ln x) k 1
皆為 associated homogeneous equation 的解
252
(3) 若 auxiliary function 在  + j 和  − j 的地方各有一個根
(未出現重根)


x cos   ln x  , x sin   ln x 
是 associated homogeneous equation 的其中二個解
(4) 若 auxiliary function 在  + j 和  − j 的地方皆有 k 個重根
x cos   ln x  , x cos   ln x  ln x, x cos   ln x  (ln x) 2 ,
,
x cos   ln x  (ln x) k 1
x sin   ln x  ,
x sin   ln x  ln x , x sin   ln x  (ln x) 2 ,
x sin   ln x  (ln x) k 1
是 associated homogeneous equation 的其中2k 個解
,
Example 4 (text page 166)
x3 y  x   5 x 2 y  x   7 xy( x)  8 y  0
auxiliary function
m  m  1 m  2   5m  m  1  7m  8  0
m3  3m2  2m  5m2  5m  7m  8  0
m3  2 m 2  4 m  8  0
 m  2   m2  4   0
253
254
4-7-5 Nonhomogeneous Case
To solve the nonhomogeneous Cauchy-Euler equation:
Method 1: (See Example 5)
(1) Find the complementary function (general solutions of the associated
homogeneous equation) from the rules on pages 244-247, 251-252.
(2) Use the method in Sec.4-6 (Variation of Parameters) to find the
particular solution.
(3) Solution = complementary function + particular solution
Method 2: See Example 6,重要
Set x = et, t = ln x
255
Example 5 (text page 166, illustration for method 1)
x 2 y  x   3xy( x)  3 y  2 x 4e x
Step 1 solution of the associated homogeneous equation
auxiliary function
m  m  1  3m  3  0
m2  3
yc  c1 x  c2 x3
Step 2-2 Particular solution W 
W1 
Step 2-3
0
x3
2 x 2e x
3x 2
u1 
x
x3
1 3x 2
 2 x3
W2 
 2 x e
W1
  x 2e x
W
m1  1
m 2  4m  3  0
5 x
u2 
x
0
1 2 x 2e x
W2
 ex
W
 2 x 3e x
Step 2-4 u1   u1dx   x e  2 xe  2e
2 x
x
256
x
u2   u2 dx  e x
Step 2-5 y p  u1 y1  u2 y2  2 x 2e x  2 xe x
Step 3
y  c1 x  c2 x3  2 x 2e x  2 xe x
Example 6 (text page 167, illustration for method 2)
x 2 y  x   xy( x)  y  ln x
Set x = et, t = ln x
dy dt dy 1 dy


dx dx dt x dt
(chain rule)
d 2 y d  dy  dt d  dy  1 d  1 dy 
  
 


2
dx
dx  dx  dx dt  dx  x dt  x dt 
1 d 2 y 1  d 1  dy  1  d 2 y dy 
 2 2  
   2  2  
x dt
x  dt x  dt  x  dt
dt 
Therefore, the original equation is changed into
d 2 y  t   2 d y (t )  y (t )  t
dt
dt 2
257
258
d 2 y  t   2 d y (t )  y (t )  t
dt
dt 2
y (t )  c1et  c2tet  t  2
y( x)  c1 x  c2 x ln x  ln x  2
(別忘了 t = ln x 要代回來)
Note 1: 以此類推
dk y 1
 k  Dt  k  1
k
dx
x
 Dt  1 Dt y
Dt means
d
dt
Note 2: 簡化計算的小技巧:結合兩種解 nonhomogeneous CauchyEuler equation 的長處
259
4-7-6 本節要注意的地方
(1) 本節公式記憶的方法:
把 Section 4-3 的 ex 改成 x,x 改成 ln(x)
把 auxiliary function 的 mn 改成 m(m  1)(m  2)
(2) 如何解 particular solution?
Variation of Parameters 的方法
(3) 解的範圍將不包括 x = 0 的地方 (Why?)
 m  n  1
260
還有很多 linear DE 沒有辦法解,怎麼辦
(1) numerical approach (Section 4-9-3)
(2) using special function (Chap. 6)
(3) Laplace transform and Fourier transform (Chaps. 7, 11, 14)
(4) 查表 (table lookup)
261
(1) 即使用了 Section 4-7 的方法,大部分的 DE還是沒有辦法解
(2) 所幸,自然界真的有不少的例子是 linear DE
甚至是 constant coefficient linear DE
262
Exercise for practice
Section 4-4
5, 6, 14, 17, 18, 24, 26, 32, 33, 39, 42
Section 4-5
2, 7, 13, 18, 31, 45, 62, 69, 70
Section 4-6
4, 5, 8, 13, 14, 17, 18, 21, 24, 25, 30
Section 4-7
11, 17, 18, 20, 21, 24, 32, 34, 36, 37, 42
Review 4
2, 19, 20, 23, 25, 26, 27, 28, 30, 31, 33, 38