@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@ÚÓ‘Ój�n€a@Î@Úœäñ€a@‚Ï‹»‹€@·rÓ:a@Âig@Ú‹©@Ü‹1a26@@ÖÜ»€a@I2@‚b«@H2013 Ibn Al-Haitham Jour. for Pure & Appl. Sci. Vol. 26 (2) 2013 On Solution of Regular Singular Initial Value Problems Luma. N. M. Tawfiq Reem. W. Hussein Dept. of Mathematics/College of Education for Pure Science( Ibn Al-Haitham) /University of Baghdad Received in:22 February 2012 , Accepted in:2 July 2012 Abstract This paper devoted to the analysis of regular singular initial value problems for ordinary differential equations with a singularity of the first kind , we propose semi - analytic technique using two point osculatory interpolation to construct polynomial solution, and discussion behavior of the solution in the neighborhood of the regular singular points and its numerical approximation, two examples are presented to demonstrate the applicability and efficiency of the methods. Finally , we discuss behavior of the solution in the neighborhood of the singularity point which appears to perform satisfactorily for singular problems. Kay ward : ODE, IVP ,Singular initial value problems. 257 | Mathematics @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@ÚÓ‘Ój�n€a@Î@Úœäñ€a@‚Ï‹»‹€@·rÓ:a@Âig@Ú‹©@Ü‹1a26@@ÖÜ»€a@I2@‚b«@H2013 Ibn Al-Haitham Jour. for Pure & Appl. Sci. Vol. 26 (2) 2013 Introduction In the study of nonlinear phenomena in physics, engineering and other sciences, many mathematical models lead to singular initial value problems (SIVP) associated with nonlinear second order ordinary differential equations (ODE) [1]. In mathematics, a singularity is in general a point at which a given mathematical object is not defined, or a point of an exceptional set where it fails to be well-behaved in some particular way, such as many problems in varied fields as thermodynamics, electrostatics, physics, and statistics give rise to ordinary differential equations of the form : y″ = f(t, y, y') , a < x < b (1) On some interval of the real line with some initial conditions. An IVP associated to the second order differential equation (1) is singular if one of the following situations occurs : a and/or b are infinite; f is unbounded at some x0∈ [0,1] or f is unbounded at some particular value of y or y′ [1] . How to solve a linear ODE of the form : A(x)y′′ + B(x)y′ + C(x)y = 0 (2) The first thing we do is, rewrite the ODE as : y′′ + P(x)y′ + Q(x)y = 0 (3) where, of course, P(x) = B(x) / A(x) , and Q(x) = C(x) / A(x) . there are two types of a point x0 ∈ [0,1] : Ordinary Point and Singular Point. Also, there are two types of Singular Point : Regular and Irregular Points, A function y(x) is analytic at x0 if it has a power series expansion at x0 that converges to y(x) on an open interval containing x0 .A point x0 is an ordinary point of the ODE (3), if the functions P(x) and Q(x) are analytic at x0. Otherwise x0 is a singular point of the ODE, i.e. P(x) = P0 + P1(x-x0) + P2(x-x0)2 +…….. = i i i xxp )( 0 0 −∑ ∞ = (4) Q(x) = Q0 + Q1(x-x0) + Q2(x-x0)2 +………. = i i i xxq )( 0 0 −∑ ∞ = (5) If A , B and C are polynomials then a point x0 such that A(x0) ≠ 0 is an ordinary point. On the other hand if P(x) or Q(x) are not analytic at x0 then x0 is said to be a singular. A singular point x0 of the ODE (3) is a regular singular point of the ODE if the functions x P(x) and x2 Q(x) are analytic at x0. Otherwise x0 is an irregular singular point of the ODE [2] . Shampine in [3] gave other definition, which illustrated by the following : If limx→x0 (x – x0)P(x) finite and limx→x0 (x – x0)2 Q(x) finite (6) that is, if both (x –x0)P(x) and (x –x0)2Q(x) possess a Taylor series at x0, then x0 is called a regular singular point, otherwise x0 is an irregular singular point . If A, B and C are polynomials and suppose A(x0) = 0, then x0 is a regular singular point if : limx → x0 (x - x0)( B/ A) and limx → x0 (x - x0)2 (C/A) are finite (7) Now, we state the following theorem without proof which gives us a useful way of testing if a singular point is regular. Theorem 1 [4] If the limx→0 P(x) and limx→0 Q(x) exist, are finite, and are not 0 then x = 0 is a regular singular point. If both limits are 0, then x = 0 may be a regular singular point or an ordinary point. If either limit fails to exist or is ±∞ then x = 0 is an irregular singular point . There are four kinds of singularities : • The first kind is the singularity at the first end point of the interval [0,1] ; i.e. , x = 0 . 258 | Mathematics http://en.wikipedia.org/wiki/Mathematics http://en.wikipedia.org/wiki/Set_(mathematics) http://en.wikipedia.org/wiki/Well-behaved @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@ÚÓ‘Ój�n€a@Î@Úœäñ€a@‚Ï‹»‹€@·rÓ:a@Âig@Ú‹©@Ü‹1a26@@ÖÜ»€a@I2@‚b«@H2013 Ibn Al-Haitham Jour. for Pure & Appl. Sci. Vol. 26 (2) 2013 • The second kind is the singularity at both ends of the interval [0,1] • The third kind is the case of a singularity in the interior of the interval; • The forth and final kind is simply treating the case of a regular differential equation on an infinite interval. In this paper, we focus on the first kind . Solution of Second Order SIVP In this section we suggest semi analytic technique to solve second order SIVP as following, we consider the SIVP : xm y''+ f( x, y, y' ) = 0 (8a) y(0) = A , y'(0) = B (8b) where f are in general nonlinear functions of their arguments . The simple idea behind the use of two-point polynomials is to replace y(x) in problem (8), or an alternative formulation of it, by osculator interpolation polynomials of order 2n+1, P2n+1 which enables any unknown boundary values or derivatives of y(x) to be computed . The first step therefore is to construct the P2n+1 , to do this we need the Taylor coefficients of y (x) at x = 0 : y = a 0 + a 1 x + ∑ ∞ =2i a i x i (9) where y(0)= a0 ,y'(0)= a1 ,y"(0) / 2! =a2 ,…, y(i)(0) / i! = ai , i= 3, 4,… then insert the series forms (9) into (8a) and put x = 0 and equate coefficients of powers of x to obtain a2 . Also we need Taylor coefficient of y(x) about x = 1 : y = b 0 + b 1 (x-1) + ∑ ∞ =2i b i (x-1) i (10) where y(1) = b0 , y'(1) =b1 , y"(1) / 2! =b2 ,…, y(i)(1) / i! =bi , i = 3,4,… then insert the series form (10) into (8a) and put x = 1 and equate the coefficients of powers of (x-1) to obtain b2 ,then derive equation (8a) with respect to x to obtain new form of equation say (11) : xm y''' + m xm y'' + df( x, y, y' )/dx = 0 (11) then, insert the series form (9) into (11) and put x = 0 and equate the coefficients of powers of x to obtain a3 ,then insert the series form (10) into (11) and put x = 1 and equate the coefficients of powers of (x-1) to obtain b3 , now iterate the above process many times to obtain a4 ,b4 ,then a5 ,b5 and so on, that is ,we can get ai and bi for all i ≥ 2, the resulting equations can be solved using MATLAB to obtain ai and bi for all i ≥ 2 , the notation implies that the coefficients depend only on the indicated unknowns a0 , a1, b0 , b1, and we get a0 , a1, by the initial condition .Now, we can construct a P2n+1(x) from these coefficients ( aisۥ and bisۥ ) by the following : P2n+1 = ∑ = n i 0 { ai Qi(x) + (-1)i bi Qi(1-x) } (12) where ( x j / j!)(1-x) 1+n ∑ − = jn s 0       + s sn xs = Q j (x) / j! we see that (12) have only two unknowns b0 and b1 to find this, we integrate equation (8a) on [0, x] to obtain : xmy'(x) – mxm-1y(x) + m(m–1) ∫ x 0 xm-2y(x) dx + ∫ x 0 f(x, y, y') dx = 0 (13a) and again integrate equation (13a) on [0, x] to obtain : 259 | Mathematics @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@ÚÓ‘Ój�n€a@Î@Úœäñ€a@‚Ï‹»‹€@·rÓ:a@Âig@Ú‹©@Ü‹1a26@@ÖÜ»€a@I2@‚b«@H2013 Ibn Al-Haitham Jour. for Pure & Appl. Sci. Vol. 26 (2) 2013 xmy(x) –2m ∫ x 0 xm-1y(x) dx +m(m-1) ∫ x 0 (1-x)xm-2y(x)dx+∫ x 0 (1-x)f(x,y,y') = 0 , (13b) Putting x = 1 in (13), then gives : b1 – mb0 + m(m-1) ∫ 1 0 xm-2 y(x) dx + ∫ 1 0 f(x, y, y') dx = 0 (14a) and b0–2m∫ 1 0 xm-1y(x)dx +m(m-1)∫ 1 0 (1-x)xm-2 y(x) dx + ∫ 1 0 (1-x)f(x, y, y')dx = 0 (14b) Use P2n+1 as a replacement of y(x) in ( 14 ) and substitute the initial conditions (8b) in (14) then, we have only two unknown coefficients b1, b0 and two equations (14) so, we can find b1, b0 for any n by solving this system of algebraic equations using MATLAB, so insert b0 and b1 into (12) , thus (12) represents the solution of (8) . Extensive computations have shown that this generally provides a more accurate polynomial representation for a given n . Examples In this section, many examples will be given to illustrate the efficiency, accuracy , implementation and utility of the suggested method. The bvp4c solver of MATLAB has been modified accordingly so that it can solve some class of SIVP as effectively as it previously solved nonsingular IVP. Example 1 Consider the following SIVP : y" + (2/x) y' – 10 y = 12 – 10 x4 , 0 ≤ x ≤ 1 I.C. y(0) = 0 , y'(0) = 0 , Exact solution is y(x) = 2 x2 + x4 It is clear that x = 0, is regular singular point and it is singularity of first kind . Now, we solve this example using semi - analytic technique , From equations (12) we have : P5(x) = x4 + 2x2 Olga [6] solve this example using Modification A domian Decomposition method and gives the exact solution . Example 2 Consider the following SIVP : y" + (2/x) y' + y = 0 , 0 ≤ x ≤ 1 with IC : y(0) = 1 , y'(0) = 0 . It is clear that x = 0 , is regular singular point and it is singularity of first kind and the exact solution is y(x) = sin(x) / x [7]. Now, we solve this example using semi-analytic technique ,From equation (12) we have : if n = 6 ,we get P13 as follows : P13 = – 0.0000000000051205 x13 + 0.0000000001756046 x12 – 0.0000000000246766464 x11 – 0.0000000250276639921x10– 0.0000000000145659x9+ 0.00000275573675x8 – 0.00000000000068686 x7 – 0.00019841269841336 x6 + 0.008333333333333 x4 – 0.1666666666678793 x2 + 1 For more details ,table (1) gives the results of different nodes in the domain, for n= 6. Also, figure(1) illustrated the accuracy of suggested method for n=6. 260 | Mathematics @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@ÚÓ‘Ój�n€a@Î@Úœäñ€a@‚Ï‹»‹€@·rÓ:a@Âig@Ú‹©@Ü‹1a26@@ÖÜ»€a@I2@‚b«@H2013 Ibn Al-Haitham Jour. for Pure & Appl. Sci. Vol. 26 (2) 2013 Ramos[7] solved this example by Linearization techniques and the results gave in figure 2 and the absolute error E(y) = │yexact – yapp.│ gave in figure 3 for h = 0.01 and 0.001 respectively . Batiha[8] solved this example by Variational Iteration Method (VIM) and the absolute error between the 5-iterate of VIM and the exact solution gave in table 2. Behavior of the solution in the neighborhood of the singularity x= 0 Our main concern in this section will be the study of the behavior of the solution in the neighborhood of singular point . Consider the following SIVP : y′′(x) + ((N − 1) / x ) y′(x) = f(y) , N ≥ 1 , 0 < x < 1 (15) y(0) = y0 , limx→0+ x y′(x) = 0 (16) where f(y) is continuous function . As the same manner in [9], let us look for a solution of this problem in the form : y(x) = y0 − C xk (1 + o(1)) (17) y′(x) = − C k xk−1(1 + o(1)) y′′(x) = − C k (k − 1) xk−2(1 + o(1)) , x → 0+ where C is a positive constant and k > 1. If we substitute (17) in (15) we obtain : C = (1/ k) (f(y0) /N )k−1 (18) In order to improve representation (17) we perform the variable substitution : y(x) = y0 − C xk (1 + g(x)) (19) we easily obtain the following result which is similar to the results in [9]. Theorem 2 [9] For each y0 > 0, problem (15), (16) has, in the neighborhood of x = 0, a unique solution that can be represented by : y(x, y0) = y0 − C xk (1 + g xk + o(xk) ) , where k, C and g are given by (18) and (19), respectively. We see that these results are in good agreement with the ones obtained by the method in [9], they are also consistent with the results presented in [10]. In order to estimate the convergence order of the suggested method at x = 0, we have carried out several experiments with different values of n and used the formula : cy0 = − log2 ( |y0n3 − y0n2| / |y0n2− y0n1 | ) (20) where y0ni is the approximate value of y0 obtained with ni ,ni = 1,2, 3, 4,… References 1. Robert, L.B. and Courtney, S. C.,( 1996) , " Differential Equations A Modeling perspective " , United States of America . 2. Rachůnková, I., Staněk, S., and Tvrdý , M., (2008) , " Solvability of Nonlinear Singular Problems for Ordinary Differential Equations " New York , USA . 3. Shampine, L. F., Kierzenka, J. and Reichelt, M. W., (2000) , " Solving Boundary Value Problems for Ordinary Differential Equations in Matlab with bvp4c " . 4. Howell , K.B., (2009) ," Ordinary Differential Equations" , USA , Spring. 5. Burden, L. R. and Faires, J. D., (2001) ," Numerical Analysis " , Seventh Edition. 6. Olga ,F., Zdeněk, Š., (2010), A Domian Decomposition Method 7. For Certain singular initial-value problems, Journal of Applied Mathematics (3): 2.91-98 . 8. Ramos, J.I. , (2005), Linearization techniques for singular initial-value problems of ordinary differential equations , Applied Mathematics and Computation 161: 525–542 9. Batiha, B. , (2009) , ( Numerical Solution of a Class of Singular Second-Order IVPs by Variational Iteration Method , Int. Journal of Math. Analysis(3): 40. 1953 – 1968. 261 | Mathematics @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@ÚÓ‘Ój�n€a@Î@Úœäñ€a@‚Ï‹»‹€@·rÓ:a@Âig@Ú‹©@Ü‹1a26@@ÖÜ»€a@I2@‚b«@H2013 Ibn Al-Haitham Jour. for Pure & Appl. Sci. Vol. 26 (2) 2013 10. Morgado, L. and Lima, P.,(2009) , " Numerical methods for a singular boundary value problem with application to a heat conduction model in the human head", Proceedings of the International Conference on Computational and Mathematical Methods in Science and Engineering, CMMSE . 11. Abukhaled, M. , Khuri, S.A. and Sayfy, A. , (2011), "A NUMERICAL APPROACH FOR SOLVING A CLASS OF SINGULAR BOUNDARY VALUE PROBLEMS ARISING IN PHYSIOLOGY ", INTERNATIONAL JOURNAL OF NUMERICAL ANALYSIS AND MODELING (8):2.353–363. Table 1: The result of the method for n = 6 of example 2 0.841470984807897 b0 0.301168678939757 - b1 Error | y(x) – P13 | P13 Exact solution y(x) xi 0 1 1 0 e-016220446049250312. 0.998334166468282 0.998334166468282 0.1 e-016 40892098500626 4 .4 0.993346653975307 0.993346653975306 0.2 0 0.985067355537799 0.985067355537799 0.3 1.110223024625157e-016 0.973545855771626 0.973545855771626 0.4 e-01611022302462516.1 0.958851077208406 0.958851077208406 0.5 0 0.941070788991726 0.941070788991726 0.6 0 0.920310981768130 0.920310981768130 0.7 e-016 22044604925031 2. 0.896695113624404 0.896695113624403 0.8 . 33066907387547 e-016 3 0.870363232919426 0.870363232919426 0.9 0 0.841470984807897 0.841470984807897 1 e-031 4.314083075427408 SSE Table 3: The comparison of the suggested method and VIM [8] of example 2 Absolute Error of VIM Exact solution y(x) xi 0 1 0 1.00000E-20 0.998334166468282 0.1 6.70000E-19 0.993346653975306 0.2 8.53100E-17 0.985067355537799 0.3 2.69223E-15 0.973545855771626 0.4 3.91600E-14 0.958851077208406 0.5 3.48972E-13 0.941070788991726 0.6 2.21760E-12 0.920310981768130 0.7 1.10021E-11 0.896695113624403 0.8 4.51811E-11 0.870363232919426 0.9 1.59829E-10 0.841470984807897 1 2.771272831459776 e-020 SSE 262 | Mathematics @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@ÚÓ‘Ój�n€a@Î@Úœäñ€a@‚Ï‹»‹€@·rÓ:a@Âig@Ú‹©@Ü‹1a26@@ÖÜ»€a@I2@‚b«@H2013 Ibn Al-Haitham Jour. for Pure & Appl. Sci. Vol. 26 (2) 2013 Figure 1: illustrate suggested method for n= 6,i.e.,P13 of example 2 . Figure 2: Linearization techniques for example 2 gave in [7] Figure 3: absolute error of method gave in [7] for h=0.01 and 0.001 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0.84 0.86 0.88 0.9 0.92 0.94 0.96 0.98 1 1.02 x y the soulution at n=6 exact p13 263 | Mathematics @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@ÚÓ‘Ój�n€a@Î@Úœäñ€a@‚Ï‹»‹€@·rÓ:a@Âig@Ú‹©@Ü‹1a26@@ÖÜ»€a@I2@‚b«@H2013 Ibn Al-Haitham Jour. for Pure & Appl. Sci. Vol. 26 (2) 2013 النظامیة الشاذة ةدائیاالبتحول حل مسائل القیم لمى ناجي محمد توفیق حسینریم ولید ) / جامعة بغدادابن الھیثم( للعلوم الصرفة كلیة التربیة الریاضیات /قسم علوم 2012تموز 2قبل البحث في: ، 2012شباط 22استلم البحث في: الخالصة للمعادالت التفاضلیة االعتیادیة الھدف من ھذا البحث عرض دراسة تحلیلیة لمسائل القیم االبتدائیة النظامیة الشاذة بوصفھا النقطتین للحصول على الحل يالتقنیة شبھ التحلیلیة باستخدام االندراج التماسي ذننا نقترح إ ،إذوبأنواع مختلفة الحل و أخیرا ناقشنا سلوك .وسھولة أداء الطریقة المقترحة ، ة یالكفاوحدود ومناقشة عدد من األمثلة لتوضیح الدقة ، متعددة و اقترحنا صیغة جدیدة مطورة لتخمین الخطأ تساعد في تقلیل . و إیجاد الحل التقریبي لھا ، في جوار النقاط الشاذة الحسابات العملیة وإظھار النتائج بشكل مرضي فیما یخص المسائل الشاذة . ، مسائل القیم االبتدائیة الشاذة المعادالت التفاضلیة االعتیادیة، مسائل القیم االبتدائیة: الكلمات المفتاحیة 264 | Mathematics