1: \documentclass[12pt,fleqn]{article}
2: \usepackage{graphicx}
3: \begin{document}
4: \title{\bf A Few Exactly Solvable Models For Riccati Type Equations}
5: \author{\bf B G Sidharth and B S Lakshmi\\Center For Applicable Mathematics and
6: Computer Sciences,\\B M Birla Science Center,\\\small Adarsh
7: Nagar, Hyderabad-500063, India}
8: \date{}
9: \maketitle
10: \begin{abstract}
11: We consider the Ricatti equation in the context of population
12: dynamics,quantum scattering and a more general context.We examine
13: some exactly solvable cases of real life interest.
14: \end{abstract}
15: \section{\bf Introduction}
16: The Riccati Equation \cite{weig,hart,ross}
17: \begin{equation}
18: y' = p(t) + q(t) y + r(t) y^2 \end{equation} appears in several
19: branches of applicable mathematics, for example population
20: dynamics and mathematical physics, as in quantum scattering.It is
21: known that it exhibits chaotic behavior \cite{cook}.It is also well known that,
22: it reduces to a second order linear differential equation by the substitution,
23: $y = \frac{u'}{u} $,\\
24: when we get, \begin{equation} u'' + g(t)u' + f(t) u = 0
25: \end{equation} (1) and (2) have been studied in detail for a long
26: time\cite{reid,sid,lak}.\\It may be mentioned that if two solutions of the
27: Riccati equation are equal at a point, then they coincide.The
28: reason lies in the continuity of the logarithmic derivative
29: $\frac{u'}{u}$ given in the substitution that lead from (1) to
30: (2).\\We will now consider exact solutions of the Riccati equation
31: in two different contexts,from the field of population dynamics
32: and the field of quantum scattering.
33: \input epsf
34: \section{Some exact solutions}
35: \subsection{A Problem In Population Dynamics}
36: We now examine coupled Riccati type equations of relevance to a population
37: model.
38: \subsubsection*{Gause's Model}
39: Consider two-species populations occurring together,and assume
40: that the growth of each is inhibited by members ,both of its own
41: and of the other species.Denoting the number of individuals in
42: species \rm{1} as $N_1$ and species \rm{2} as $N_2$,we have the Gause's
43: competition equations:\cite{pie}
44: \begin{equation}\frac{1}{N_1}
45: \frac{dN_1}{dt} = r_1- a_1 N_1 - a_2 N_2\end{equation}
46: \begin{equation}\frac{1}{N_2} \frac{dN_2}{dt} = r_2-a_3 N_1 - a_4 N_2
47: \end{equation}
48: where $r_{1}$,$r_{2}$,$a_{1}$,$a_{2}$ are defined below. Thus we
49: are assuming that the per capita growth of each population at an
50: instant is a linear function of the sizes of the two competing
51: populations at that instant.Each population would grow
52: logistically if it were alone with logistic parameters $r_1$ and
53: $a_{1}$ for species \rm{1} and $r_{2}$ and $a_{2}$ for species
54: \rm{2} .\\ In general the simultaneous differential equations
55: cannot be explicitly solved.We now consider a particular set of
56: circumstances in which they can be solved.Thus for (3) and (4) we
57: specialize to, with a more convenient notation ,
58: \begin{equation}
59: \frac{dy}{dt}= 1-xy^{2}
60: \end{equation}
61: \begin{equation}
62: \frac{dx}{dt}= 1+yx^{2}
63: \end{equation}
64: From (5) and (6) we get,
65: \begin{equation}
66: \frac{dy}{dx}= \frac{1-xy^{2}}{1+yx^{2}}
67: \end{equation}
68: Integrating (7) we have
69: \begin{equation}
70: y-x = -x^{2}y^{2}+ C
71: \end{equation}
72: From (8) when $y = 0$,$C=-x_{0}$ so that $y < x$ \\Reverting back
73: to $N_{1}$ and $N_{2}$, this is \begin{equation} N_{1}-N_{2} =
74: -{N_{2}}^2{N_{1}}^2+ C \end{equation} From (9) when
75: $N_{1}=0$,$C=-N_{0}$ so that $N_{1} < N_{2}$.\\The Figure below
76: illustrates a particular case of the
77: above solution with the two populations $N_{1}$ and $N_{2}$ along Y and X axes.\\
78:
79: \includegraphics[bb= 0 0 224 138]{potent.eps}
80:
81: \subsection{Problems from Scattering Theory}
82: By reversing the method used above ,the quantum mechanical radial
83: Schr\"{o}dinger equation
84: \begin{equation}
85: u'' - H(r) u = 0
86: \end{equation}
87: can be reduced to a Riccati equation
88: \begin{equation}
89: v' + v^{2}= H(r)
90: \end{equation}
91: by the substitution $v = \frac{u'}{u}$. The form (10)or (11) is
92: used in phase shift analysis for example in Calogero's variable
93: phase approach \cite{sidh}.It is also possible to use the form
94: (11) for building up an iterative procedure\cite{cal} .We would
95: now like to point out that (11) can be used in a different
96: context for providing exact solutions for specified classes of the
97: potential function f(r). Let us write in \nolinebreak (11) \[v
98: \equiv \frac{1}{f}\] \[ f'=1-gf
99: \] Whence we get
100: \begin{equation}
101: H= \frac{g}{f}=\frac{ge^{\int{g}dr}}{\int{e^{\int{g}dr}+C}}
102: \end{equation}
103: 1.We put g=$\alpha$ in (12) \[ H=\frac{\alpha^{2}}
104: {1+c\alpha e^{-\alpha r}}\]
105: The potential H is now of
106: the form \[
107: H =\frac{D}{A+Be^{\alpha r}}
108: \]
109: This is the well-known Wood-Saxon potential. The solution u is
110: given by
111: \[
112: u=C_{1}(e^{\alpha r}+C_{2}\alpha)
113: \]
114: 2.Putting g=r
115: in (12), the potential H is given by
116: \[ H= \frac{r e^{r^2/2}}{\int e^{r^2/2} dr +
117: c_{1}}\]
118: This is a modified Gaussian potential and the solution u is given by
119: \[
120: u = K \int e^{-r^2}dr + C_{3}
121: \]
122: The Graph is shown in the figure
123: below and is seen to fall very steeply indicative of a confined state or
124: particle.\\ \vspace{9 pt} \\
125: \includegraphics[bb= 0 0 224 138]{ricc.eps}
126:
127: 3.Putting g=1/r in
128: (11),the potential H is given by
129: \[ H= \frac{2}{r^2+c_{4}}
130: \]
131: This is a shifted inverse square potential and the
132: solution u is given by
133: \[
134: u= K'(r^2 +C_{4})
135: \]
136: \section{Other Exact and Asymptotic Solutions}
137: We now consider some asymptotic solutions of (11),which we write as,
138: \begin{equation}
139: u' = u^{2} + f \end{equation}
140: Let \[ \int{f} dr \equiv {g}, \]
141: f and g being bounded functions.
142: So (13) can be written as ,
143: \begin{equation} z' = (z+g)^2 > 0, z=u-g, \end{equation}
144: which shows that z is an increasing function of r.Suppose z is
145: unbounded.So for large r,we should have from (14)
146: \[ z \approx z^2 \]
147: whence \[ z = \frac{1}{c-r} \] which $ \rightarrow 0 $ as $ r
148: \rightarrow \infty $.\\ This is a contradiction. Therefore z is
149: bounded and so also u ,that is $ u\longrightarrow M
150: $\quad as \quad $r \longrightarrow \infty $\\
151: Therefore $u'\approx 0 $ (asymptotically)\\
152: Therefore for large r,(13) becomes
153: \[ u^{2} = -f, \textup{whence} \]
154: \begin{equation} u = \pm (-f)^{1/2} \end{equation}
155: By way of verification of (15),let us consider (13) with,
156: \[ f = - \frac{\alpha^2}{r^{2}}\]
157: So we expect that for large r, $u \sim$ $ \pm$ $ \frac{\alpha}{r}$
158: by (15). Let us put \begin{equation} u = \frac{\beta}{r}
159: \end{equation} in (13). So \[-\frac{\beta}{r^{2}} =
160: \frac{\beta^{2}}{r^{2}} -\frac{\alpha{2}}{r^{2}} \] or \[ \beta =
161: \frac{-1 \pm\sqrt{1+4\alpha^{2}}}{2} \approx \pm
162: \alpha\hspace{3mm} \textup{if} \hspace{3mm}\alpha \ \gg 1 \]
163: \textup{That is} \[ u = \pm
164: \frac{\alpha}{r} \] (everywhere, and so also for large r). Using
165: this example,with transformations of the independent variable,we
166: can generate similar solutions. For example, if we substitute for
167: r, $ t = t(r)$, (13) becomes
168: \[ t'\dot{u} = u^2 + f,\]
169: where the dot denotes the derivative with respect to t.The choice
170: $t'=f(r)$, leads to a similar equation, and one can verify that
171: for
172: \[f(r) = \frac{a}{3} r^{-2/3} - a^2 r^{2/3},\hspace{5mm} u = a
173: r^{1/3}\] is a solution.More generally as can be easily verified
174: $a r^n$ is a solution for $ f(r)=n a r ^{n-1}-a^2 r^{2n}$ and so
175: on.However in these examples,
176: neither u nor u' are asymptotically bounded.\\
177: Finally it maybe observed that if one solution of the Riccati equation
178: (13) is known,then others could be derived therefrom\cite{ross}.
179: \begin{thebibliography}{99}
180: \bibitem{weig}Weigelhofer,W.S.and Lindsay, K.A., ``Ordinary Differential Equations And Applications'',Horwood Publishing Limited,Chichester,1999,p.8 ff
181: \bibitem{hart}Hartman,P.,``Ordinary\hspace{1mm}Differential\hspace{1mm}Equations''Birkhauser,Boston,1982,p.331
182: \bibitem{ross}Ross,S.L.,``Differential\hspace{1mm}Equations'',Blaisdell\hspace{1mm}Publishing\hspace{1mm}Company, Massachussets,Toronto,London 1964
183: \bibitem{cook}Cook,A.,``The Observational Foundations Of Physics'',Cambridge University Press,Cambridge,1994,p.73
184: \bibitem{reid}Reid,J.L.,Physics Letters,97A(8)1983,pp.309-310
185: \bibitem{sid}Sidharth,B.G.,\hspace{1mm}and\hspace{1mm}Maqbool,Indian\hspace{1mm}Journal\hspace{1mm}of\hspace{1mm}Chemical\hspace{1mm}Sciences,Vol.\textrm{1},1987,pp.31-36
186: \bibitem{lak}Lakshmi,B.S.,Chaos,Solitons and Fractals,16-2003,183-186
187: \bibitem{pie}Pielou,E.C.,``Mathematical Ecology'',John Wiley and Sons, New York,1977
188: \bibitem{sidh}Sidharth,B.G.,J.Math.Phys.,24,878(1983)
189: \bibitem{cal}Calogero,F.,``Variable\hspace{1mm}Phase\hspace{1mm}Approach\hspace{1mm}to\hspace{1mm}Potential\hspace{1mm}Scattering,Academic Press,New York,1967,pp.31-36
190: \end{thebibliography}
191: \end{document}
192: