2.22.1 Algorithm
ode internal name "first_order_ode_chini"
This ode is normally generated when we get an Abel ode of first kind \(f_{0}+f_{1}y+f_{2}y^{2}+f_{3}y^{3}\) and then remove
the square term \(f_{2}\) using the transformation \(y=u\left ( x\right ) -\frac {f_{2}}{3f_{3}}\). Again as mentioned above, this is done when
the Abel invariant is constant. See above section.
Now we check if the Chini invariant is also constant or not. The Chini invariant is given by
\[ \Delta =f^{-n-1}h^{-2n+1}\left ( fh^{\prime }-f^{\prime }h-ngfh\right ) ^{n}n^{-n}\]
And if this comes out to be constant (i.e. do not depend on
\(x\)), then we can now solve the
Chini ode using method given in Kamke page 303.
Otherwise there is no general method to solve it. This below is my translation of Kamke
1.55, page 303 on Chini ode. He says, given ode
\begin{equation} y^{\prime }=f\left ( x\right ) y^{n}+g\left ( x\right ) y+h\left ( x\right ) \tag {1}\end{equation}
If for a suitable constants
\(\alpha ,\beta \)\begin{equation} \left ( \frac {h}{f}\right ) ^{\frac {1}{n}}=e^{\int gdx}\left ( \beta +\alpha \int he^{-\int gdx}dx\right ) \tag {2}\end{equation}
Let
\begin{equation} z=\left ( \frac {h}{f}\right ) ^{\frac {1}{n}} \tag {3}\end{equation}
Then
\[ \alpha =\frac {z^{\prime }-gz}{h}\]
We
get the solution of the original ode as
\begin{align} y & =\left ( \frac {h}{f}\right ) ^{\frac {1}{n}}u\left ( x\right ) \nonumber \\ u & =y\left ( \frac {h}{f}\right ) ^{\frac {-1}{n}} \tag {5}\end{align}
The book here was not clear. But it seems we have two cases. If \(\Delta =0\) then solution is
\begin{equation} \int \frac {du}{u^{n}-\alpha u+1}+c_{1}-\int \left ( \frac {h}{f}\right ) ^{\frac {1}{n}}hdx=0 \tag {6}\end{equation}
And if
\(\Delta \neq 0\)
then solution is
\begin{equation} \int \frac {du}{\frac {u^{n}}{\Delta }-u+1}+c_{1}-\alpha \int \left ( \frac {h}{f}\right ) ^{\frac {1}{n}}hdx=0 \tag {7}\end{equation}
For
\(h=0\) the ode is Bernoulli. Lets try to figure how the above works on
number of examples. In the above
\(\Delta \) is Chini invariant. The book also was not clear what
\(\beta \)
is. It seems it is always zero. This is what is done in all examples below and solutions
found were verified correct by Maple.