first order o.d.e.

LINEAR O.D.E. LINEAR O.D.E. WITH CONSTANT COEFFICIENTS

P(x)dx
y ¢ + P( x)y = Q(x ) integrating factor m ( x ) = eÚ y ¢ + ay = Q(x ) integrating factor m (x ) = e
ax

general general
solution y (x) = cm -1 (x ) + m-1 (x ) Ú m(x)Q(x )dx solution y (x) = ce-ax + e-ax Ú eax Q(x)dx

Initial Value Problem y (x 0 ) = y 0 Initial Value Problem y (x 0 ) = y 0

† †
x x
a ( x 0 -x )
solution y (x) = y 0m(x0 )m-1 (x) + m-1 (x) Ú m(x)Q(x)dx solution y (x) = y 0 e + e-ax Ú e Q(x)dx
ax

x0 x0

† † standard
M (x,y )dx + N(x,y )dy = 0 differential
NONLINEAR EQUATIONS form

equation is separable if
EXACT SEPARABLE it can be represented by M1 (x)M 2 (y )dx +N1 (x)N2 (y )dy = 0
†
test for
exactness ∂M ∂N then
if =
∂y ∂x then variables can
be separated f (x )†
dx + g(y )dy = 0
there exists function
df(x, y) = M (x, y)dx + N(x, y)dy
f (x, y) such that

general È ∂ ˘
general
solution Ú f (x )dx + Ú g( y)dy = c
solution f(x,y ) = Ú Mdx + Ú ÍÎN - ∂y Ú Mdx˙˚dy = c
or
È ∂ ˘ check for suppressed solutions
f(x,y ) = Ú Ndy + Ú ÍÎM - ∂x Ú Ndy˙˚dx = c
†
reduce replace M by mM
to exact › and N by mN › reduce to
separable

† equation is homogeneous
INTEGRATING FACTOR
equation becomes exact when
HOMOGENEOUS if coefficients are M (lx,ly ) = lmM(x,y )
multiplied by intgrating factor m
homogeneous functions
of the same order N(lx,ly ) = lmN(x,y )
∂M ∂N homogeneous equation can be reduced to separable by:
-
∂y ∂x Ú f ( x )dx †
a) if
N
= f(x) fi integrating factor m(x) = e a) change of variable y = ux dy = udx + xdu or x = vy dx = vdy + ydv
†
b) conversion to
∂M ∂N polar coordinates x = r cosq y = r sinq dx = cos qdr - r sin qdq
- + dy = sinqdr + r cos qdq
∂y ∂x † Ú g( y )dy † †
†
b) if = g(y ) fi integrating factor m(y ) = e homogeneous equation can 1
M be reduced to exact by m=
integrating factor:
† N1 (x)M 2 (y ) †
†
SPECIAL EQUATIONS †
†
†BERNOULLI RICATTI CLAIRAUT
†
n 2
y ¢ + P( x)y = Q(x )y y ¢ = P( x)y + Q( x)y + R (x ) y = xy ¢ + f (y ¢)

change of
if one particular solution is known u1 then
1
variable y = z 1-n reduces equation to change of 1 set of
variable y = u1 + leads to solutions y = c x + f (c )
z
linear
o.d.e. z¢ + (1- n)P(x)z = (1- n)Q(x )
linear
o.d.e. z¢ + (2Pu1 + Q )z = -P
† x = -f ¢( t)
and one
† of parametric
back substitution z = y 1-n
change
variable y=-
w¢
leads to solution y = f( t) - tf ¢( t )
Pw
†
linear Ê P¢ ˆ
y=0 is also a solution for n>0 2nd order w¢¢ - Á + Q ˜w¢ + RPw = 0
o.d.e. ËP ¯
†
†
APPROXIMATE METHODS

IVP PICARD'S METHOD † TAYLOR SERIES y (x 0 ) = y 0

y ¢ = f (x, y) choose starting approximation y (0 ) ( x) then
y ¢( x 0) = f( x0 , y0 )
x y ¢(x0 ) y¢¢(x 0 ) 2
y (x) = y (x 0 ) + (x - x ) + (x - x ) + ...
y (x 0 ) = y 0 iterate y ( j+1) (x) = y 0 + Ú f(x,y( ) (x))dx
j 1
0
2!
0
y ¢¢(x 0 ) =
∂
f(x 0 ,y 0 )
x0 ∂x
where j=0,1,2,... is the number of iteration
...

†
† †