Classification of Differential Equations for finding their solutions ( By Dr.

Akhilesh Jain, CIST, Bhopal)

Differential equations: (1) ODE, Ordinary diff. eq. (2) PDE , Partial Diff. Eq.
First order ODE Second order ODE Cauchy’s Homogeneous diff. equation
separation of Variables Second order Second order ode with d2y dy dy dy
Separate the variables y from x, i.e., by collecting on x2
 x  y  Q ( x ) Put x  ez or z  log x x   Dy
ode with const. variable coefficient dx 2 dx dx dz
one side all terms involving y together with dy, while coefficient
all terms involving x together with dx are put on the d2y
x2  D ( D  1) y and convert the diff. eq. in const. coff . equ. And apply the
other side. Integrate both sides. dx 2
previous methods
Homogeneous diff. equation: d2y dy Second order non homogeneous ODE with variable coefficient
Complete solution P Q y  R
Put y= v x and dy/dx = v+ x dv/dx in the given dx 2 dx (other than Cauchy’s equation)
equation , and use seperation of variavle method One part of Integral is given or can be find by inspection
Linear diff. equation: Y=C.F.+ P.I.
If r(x) = 0 for all x, then, the equation is said to be homogeneous. Let y= u.v is complete solution (Where u is the one part of the solution).
This method is used when the equation is in the form Substitute the values of y , y’ and y’’ in given eq..
If r(x) ≠ 0 for all x, then, the equation is said to be nonhomogeneous.
dy Now we get Second order diff. eq. with constant coefficients in terms of v.
of  p ( x) y  q ( x) where p(x) and q(x) – continuous Solving Second-Order Linear Homogeneous Differential Equations With
dx Constant Coefficients (When r(x) = 0 ): Find “auxiliary equation” by Solve this reduced equation for v and find the value of Complete solution y=u.v.
functions may or may not be constants.
d2y
Method-I : If 1+P/a +Q / a2 =0 then y= eax is one part of the solution (C.F.).
dy
p ( x ) dx Replacing with m2, with m, and y with 1. If 1+P +Q =0 then y= ex is one part of the solution (C.F.).
Solution: Find , I.F.= e Then Solution : y. dx 2 dx If 1-P +Q =0 then y= e-x is one part of the solution (C.F.).
I.F. =  I .F .Q ( x ) dx  C Case- I : If auxiliary equation has real and distinct roots m1 and m2 then Method -II
Exact Diffrential equations:A diffrential , C.F.= y  yc ( x)  c1em1x  c2em2x Method-I : If m (m-1)+ Pmx+Qx2=0 then y= xm is one part of the solution (C.F.).
If P+Qx =0 then y=x is one part of the solution.
equation obtained from its premitive(solution)directly Case-II : If auxiliary equation has real and equal root m1 = m2 =m then
If 2+2Px+Qx2=0 then y= x 2 is one part of the solution (C.F.).
by diffrentiation without any subsequent operation of C.F. = y  yc ( x)  c1emx  c2 xemx  (c1  c2 x)emx II Removal of first derivatives/ Reduction in normal form
elimination or reduction.
General form Mdx+Ndy=0, M and N are function of x Case –III : If auxiliary equation has complex roots m     i )then 1
 Pdx
2
M N x
C.F. = y  yc ( x)  e (c1 cos  x  c2 sin  x) ue We get Normal equation
Condition for exactness: 
and y. y x d 2v 1 dp 1 2
Solution of Second-Order Linear Nonhomogeneous Differential  Iv  R / u , Where I  Q   P
Equations With Constant Coefficients (When r ( x )  0 )
Solution of exact eq.: dx 2 2 dx 4

y cons tan t
Mdx  
xcons tan t
Ndy  C (write common
1 1 ax Then use previous methods.
When r(x)= eax P.I .  e ax  e III Change of independent variables
terms once ) f ( D) f (a)
dz d 2z dz dz
M N ax
When r(x)= e (Special Case When f(a)=0) Choose z such that Q1  Q / ( ) 2  c 2  1 Find P1  [ 2  P ] / ( ) 2 and
Conversion to exact form: If  then Put D=D+a and Solve the equation for 1= x0 or e0x dx dx dx dx
y x
1 eax dz 2
equation is not exact, we have to convert it in exact P.I .  eax  1 R1  R / ( ) Substitute these values in Transformed equation
form using following methods: f ( D) f ( D  a) dx
Method-I : If Mx+Ny≠0 ( a small term) , then take When r(x)= Sinax or Cosax d2y dy
1 Put D2=-a2 and Solve the equation for D by rationalization of the equation  P1 ( x )  Q1 ( x ) y  R1 ( x ) Solve this transformed equation by previous
I.F.( Integral Factor) = , and multiply the dx 2 dx
Mx  Ny (same for cosax) , Except f(-a2) ≠ 0 methods.
1 1
equation by I.F. to reduce in exact form. P.I .  sin ax  sin ax If f(-a2) = 0 then use1. Method of variation of Parameters: This method is used to find the complete
f (D2 ) f (a 2 ) solution, without finding the particular solution.( for const. coefficient and variable
Methid-II If f1(x,y)y dx + f2(x,y)x dy=0 , If Mx-Ny≠0 1 x 1 x
diff. Eq. both)
( a small term) , then take I.F.( Integral Factor) = P.I .  sin ax  cos ax , P.I .  cos ax   sin ax
1. Find the C.F. of the equation
f (D2  a2 ) 2a f (D2  a2 ) 2a 2. Let Complete solution is y= A u(x) + B v(x), where u(x) and v(x) are the parts of
1
, and multiply the equation by I.F. to reduce C.F. .Where A and B can be determined by
Mx  Ny
1 Ru Rv
in exact form. When r(x)= xm then P.I .  xm  f ( D) 1 x m A dx  c1 and B   dx  c2
f ( D)
1  M N 
u v u v
 Expand Series f(D)-1 using
Methid-III: When    f ( x) function of
N  y x  (1-x)-1= 1+x +x2+x3...... or (1+x)-1 =1-x +x2 - x3...... u' v' u' v'
When r(x)= eax V (Where V is the function of x)
x only, then I .F .  e 
f ( x ) dx ax Simultaneous Diff. equation
, multiply this I.F. to Put D= D+a for e and then use Previous formua (For solving V )
Change the equations in D- Notations and eliminate one variable using methods of
equation and make exact. 1 1
P.I .  x. e ax  e ax V eliminations
1  M N  f ( D) f ( D  a)
Methid-IV: When      f ( y ) function Later solve the equation for one variable and find solution using previous methods.
M  y x  When r(x)= x V (Where V is the function of x) Again substitute the value of one variable ( found by solution) and find the value of
1 f '( D )
of y only, then I .F.  e 
 f ( y)dy P.I .  x.V  x V For solving V use previous formula. another variable.
, multiply this I.F. to f ( D) [ f ( D )]2
equation and make exact.