Hi songoku!
songoku said:
So, differential equation will have complementary solution only if the form :
dy/dx + (a)y = r(x) ?
No, the complementary solution is where the RHS is
zero (the "homogenoeus equation").
particular integral = particular solution?
This is what I read on my book :
"the general solution of a differential equation, involving an arbitrary constant, represents more than one, even infinitely many, solutions. the solutions obtained by assigning definite value to the constant is called a particular solution. typically, the value is determined by making the solution satisfy a given condition, traditionally referred to as initial condition."
So I think particular integral is not the same as particular solution.
No, I think they're the same …
"the solution obtained by assigning a definite value to the constant"
is a particular solution (in this case, it has the constant = 0).
Remember, a "particular solution" is any, repeat ANY, solution to the complete equation.
To find particular solution, we have to find the general solution first.
No NO NOOOO!
You
can find the particular solution first.
It usually has
nothing to do with the general complementary solution (in this case, it was linear, and the complementary solution was exponential).
how to make wild but
intelligent 
, such that you can guess polynomial? and how to apply it to solve the question?
I believe there are some rules somewhere, but I don't remember them.
Usually, it's fairly obvious … you look at the RHS, and imitate it …
in this case, the RHS is a polynomial, so you try a polynomial.
I found the general solution : y = (6x+5)/3 - 2/3 + Ce^(-3x). You mean the particular integral is (6x+5)/3 - 2/3 ?
Yes,
except you'll lose marks in the exam if you don't simplify that, to 2x +1.
