Method of Frobenius.

    This method is attributed to the german mathematician Ferdinand Georg Frobenius (1849-1917 ).  Assume that  [Graphics:Images/FrobeniusSeriesProof_gr_35.gif]  is regular singular point of the differential equation

        [Graphics:Images/FrobeniusSeriesProof_gr_36.gif].  
    
A Frobenius series (generalized Laurent series) of the form  

[Graphics:Images/FrobeniusSeriesProof_gr_37.gif]

where  [Graphics:Images/FrobeniusSeriesProof_gr_38.gif]  can be used to solve the differential equation.  The parameter [Graphics:Images/FrobeniusSeriesProof_gr_39.gif] must be chosen so that when the series is substituted into the D.E. the coefficient of the smallest power of  [Graphics:Images/FrobeniusSeriesProof_gr_40.gif] is zero.  This is called the indicial equation.  Next, a recursive equation for the coefficients is obtained by setting the coefficient of  [Graphics:Images/FrobeniusSeriesProof_gr_41.gif]  equal to zero. Caveat. There are some instances when only one Frobenius solution can be constructed.

Definition (Indicial Equation).  The parameter [Graphics:Images/FrobeniusSeriesProof_gr_42.gif] in the Frobenius series is a root of the indicial equation  

        [Graphics:Images/FrobeniusSeriesProof_gr_43.gif].

Assuming that the singular point is  [Graphics:Images/FrobeniusSeriesProof_gr_44.gif],  we can calculate  [Graphics:Images/FrobeniusSeriesProof_gr_45.gif]  as follows:

        [Graphics:Images/FrobeniusSeriesProof_gr_46.gif]
and
        [Graphics:Images/FrobeniusSeriesProof_gr_47.gif]

Definition of  [Graphics:Images/FrobeniusSeriesProof_gr_48.gif]  We state the following definition of  [Graphics:Images/FrobeniusSeriesProof_gr_49.gif]

        [Graphics:Images/FrobeniusSeriesProof_gr_50.gif].  

The exponents of the singularity are the roots  [Graphics:Images/FrobeniusSeriesProof_gr_51.gif]  of  [Graphics:Images/FrobeniusSeriesProof_gr_52.gif].   

 

Derivation.

 

Starting with the differential equation  

[Graphics:../Images/FrobeniusSeriesProof_gr_53.gif]

Rewrite it in the form  

[Graphics:../Images/FrobeniusSeriesProof_gr_54.gif]

Multiply each term by the factor  [Graphics:../Images/FrobeniusSeriesProof_gr_55.gif].   

[Graphics:../Images/FrobeniusSeriesProof_gr_56.gif]


Use series for all the terms  [Graphics:../Images/FrobeniusSeriesProof_gr_57.gif].  

[Graphics:../Images/FrobeniusSeriesProof_gr_58.gif]

Make the substitutions  [Graphics:../Images/FrobeniusSeriesProof_gr_59.gif],  [Graphics:../Images/FrobeniusSeriesProof_gr_60.gif] and regroup the second and third terms in the D. E. as follows.  

 

[Graphics:../Images/FrobeniusSeriesProof_gr_61.gif]

Making all the series substitutions we get  

[Graphics:../Images/FrobeniusSeriesProof_gr_62.gif]

 

Move the terms [Graphics:../Images/FrobeniusSeriesProof_gr_63.gif] and [Graphics:../Images/FrobeniusSeriesProof_gr_64.gif]  into the summations where they belong  

[Graphics:../Images/FrobeniusSeriesProof_gr_65.gif]

 

Now look at the first term in each of the series, multiply and add as indicated to obtain the term with the lowest power of  x.  Hint. It is found by changing all of the upper limits in the summations from  [Graphics:../Images/FrobeniusSeriesProof_gr_66.gif] to [Graphics:../Images/FrobeniusSeriesProof_gr_67.gif].  

Mathematica can be of assistance for this computation by changing the upper limit in the summations from  [Graphics:../Images/FrobeniusSeriesProof_gr_68.gif] to [Graphics:../Images/FrobeniusSeriesProof_gr_69.gif].  

 

[Graphics:../Images/FrobeniusSeriesProof_gr_70.gif]

The result is

[Graphics:../Images/FrobeniusSeriesProof_gr_71.gif]

This equation is equivalient to  [Graphics:../Images/FrobeniusSeriesProof_gr_72.gif]  and since  [Graphics:../Images/FrobeniusSeriesProof_gr_73.gif]  we can cancel it and the common factor  [Graphics:../Images/FrobeniusSeriesProof_gr_74.gif]  to get

        [Graphics:../Images/FrobeniusSeriesProof_gr_75.gif].
        
We have arrived at the indicial equation which was our goal.

The roots of   [Graphics:../Images/FrobeniusSeriesProof_gr_76.gif] are denoted  [Graphics:../Images/FrobeniusSeriesProof_gr_77.gif],  and could be real or complex.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

(c) John H. Mathews 2004