[EER Derinitions Linear differential equations are those in which the dependent variable and its derivatives occur only in the first degree and are not multiplied together. Thus the general linear differential equation of the nth order is of the form Sty Tens aa Paget ages where py, Py; »-» Py and X are functions of x only. Linear differential | equations with constant eo-cfficients are of the form dy. gt STS where ky, Ray oy hy are constants. ‘Such equations are most important in the study of electro-mechanical vibra- tions and other engineering problems. (1) THEOREM Thy, Senet abe equation fy, ay # a dx? te then ey, +ey¥, (= 4) is also its solution. Since y = y, and y =y, are solutions of (1). ot gy =X tt hy eX s+ hy =0 a) ayy an a onl) ae ae ay, , ey and a +h, pe + 8) Ifc,, cy be two arbitrary constants, then AMG +692). a Mey + C292) Se tet hile, + eyy2) amt2 {a my en ae” ve T = 6((0) + 6,(0) = {By (2) and (3)) du,, a ie, Sey, shy a0 4) an a nd This proves the theorem. (2) Since the general solution of a differential equation of the nth order contains n arbitrary constants, it follows, from above, that if'y4, 924 Vg» +» J,» aren independent solutions of (1), then ey, + cg¥y + + ¢,)), (= u) is its complete solution. (8) Ify =v be any particular solution of dty., dy oop in onl a A aah te thy =X (5) then ay, tat hak (6) a Adding Girand (Oj-we have EEO) aie) +h,(u+v)= a8 This shows that y = u + v is the complete solution of (5). The part w is called the complementary function (C.F.) and the part v is called the particular integral (P-L) of (5). the complete solution (C.S.) of (5) is y = C.F. + P.I. * Thus in order to solve the question (5), we have to first find the C.F., i.e., the complete solution of (1), and then the P.L., i.e. a particular solution of (5). OPERATOR D Denoting 4, a £ ete. by D, D®, D¥ ete., so that _ de dt ds 2 @ < dy, EY = pry, a = DYy etc., the equation (5) above can be written in the symbolic form (D? + k,D*-! + we de ax a thy =X, ie, fDy=X, where f(D) =D" + AD"! +... + hy» £48 polynomial in D. ‘Thus the symbol D stands for the operation of differentiation and oan be treated much the same as an algebraic quantity ie,, f(D) ean be factorised by ordinary rules of algebra and the factors may be taken in any order. For instance #2 ooh ly ~ 3y =(D* + 2D ~ 3) y = (D+ 3(D — ly or (D- 1D + 3) y. RULES FOR FINDING THE COMPLEMENTARY FUNCTION ay op Ong ue ey eth ath Se To solve the equation +h,y=0 a) where k's are constants. ‘The equation (1) in symbolic form is (DP + h,DP1 + kDY-? +. + kyl = 0 @) Its symbolic co-efficient equated to zero ie. DP +h,D"-14 kD? +. is called the auxiliary equation (A-E.). Let m,, we Case I. If all the roots be real and different, thon (2) is equivalent to (D- my) O-m,) .. (D—m,y = 0 1B)‘Now (3) will be satisfied by the solution of (D — m,,)y = 0, i, by a —m,y=0. ‘This is a Leibnite's linear and LF, = &"™" = itssolutionis ye =c,,ie,y=¢,e""" Similarly, since the factors in (3) can be taken in any order, it will be satisfied by the solutions of (D-m,)y =0,(D— may = Dete. i.e, byy = ee", y = ene" ete. Thus the complete solution of (1) is y = e,e"* + ce" +... +¢,¢"" AA) Case IL If two roots are equal (i.e., m, = m,), then (4) becomes Yay +ehe™ +ege™ +. tee y=ce™ see +o tee" [> ¢, +, = one arbitrary constant C] It has only n ~ 1 arbitrary constants and is, therefore, not the complete solution of (1). In this case, we proceed as follows : ‘The part of the complete solution corresponding to the repeated root is the complete solution of (D —m,) (D-myy=0 Putting (D ~ m,)y =2, it becomes (D~m,)= = 0 or 42 —myz=0 ‘This is a Leibnitz’s linear in z and LF.= € its solution is ze" =c, or 2=c,e"* Thus D-myszece™ or Y —mysce™ (5) Its LF. being ¢~"™, the solution of (5) is r= fee de +e, ye (ex + cy)" eye +c, oF ‘Thus the complete solution of (1) is y =(c,x +¢,)e"* +e,e™" +... 40,0" If, however, the A.E. has three equal roots (i.e., 1m, = my = mq), then the complete solution is Ya (ep? + ex + eg) oH Hey +a eee Case IIL. If one pair of roots be imaginary, i.e.,m, = 0+ if, my = a — iB, then the complete solution is ya cet 4 eye geye™ 4.400 = eee + ce) +ege™ Fa Heye™ = e™[ey(cos Bx +i sin Bx) +c, (cos Br —i sin Bx)} +e,e"" +... 6,67 te by Euler's Theorem, e = cos @ + i sin 0) =e (C, cos fix + C, sin Bx) +56" +... +6," where C, =¢, +¢, and C, = ile,e,).. Case IV. If two points of imaginary roots be equal i.e., m, =m, = t+ iP, my =m, = 0—if, then by case II, the complete solution is y =e [(eyx + c,) cos Pe + (eg + e,) sin Br] +... +c,€7" Solution. Given equation in symbolic form is (D® + 5D + 6) x Its AB. is D?+ 5D +6 . (D+2)(D +3)=O whence D 3. s CS.ise=cet eee and SH =-me%—sog-* Whent=0,x =0. O=e+e, @ Whent=0,dx/dt=15 + 1 20, — 30 °Liz Hiawen Enciveenine Matrewarics Solving (i) and (ii), ¢, = 15, ¢, =—15. Hence the required solution is x = 15 (e-¥ ~ e~*), Example 18.2. Solve Teo +9x=0 Solution. Given equation in symbolic form is (D? + 6D + 9) = AE. isD?+6D+9=0, ie, (De ahc 0 whence = Hence the CS. is x = (c, + og). Example 13.3. Solve (D' +.D° + 4D +4)= Solution. Here the AE. is D? + D? + 4D +4 ie, (D?+4)D41 Hence the CS.is y= ¢,e* + (c, e0s 2x +0, sin 2x) ie, y= cie"* +e, cs 2x + ¢5 sin 2e. Example 134. Solve (i) (Df 4D +4) y =0 (Bhopal, 2008) (ii) (D? + DS y = Owhere D = did. Solution. (i) The A.E. equation is D'-4D?+4=0 or (D?-2)?=0 m Dt=2,2 ie,D=+ V2, V2. Hence the C.S. is (c, + ex) + (cy + ego) [Roots being repeated] Gy The AB. equation is (D? + 1)? =0 D=+i,2i,ti. Hence the C8. is y= e Ko, +02 + 639) coax +10, + egt + oge2) ein x] ie, Y=(c, He, 40,44) C08 x 4(c, + e,2 46,44) sin x, Example 13.5. Solve 12 + 4x =0. at Solution. Given equation in symbolic form is (D* + 4)x=0 . D'44=0 or (D'+4D?44)—4D?=0 or (D?+2)*-(2D)*=0 or (D* + 2D + 2)(D*— 2D +2)=0 2 either D8 +2D+2=0 or D?-2D+2-0 whence D aa ang 224° ® je,D=-14iand14i. Hence the required solution ix =e (cos +6, sin) +6! c08 +e, si). GECLTeceesl Salve 2, 1 GF - 48 4 1s =0, x0), HO -2 WLU, 2008) Le abi 6 Ay 3. 7 ¥ayF 4y'=0. ‘ 4 2% 4y20. WTU,20008) al By gdPy yay 6, a de de ae 6 feel #18y =0. © WEN-T-U., 2005) 7. (4D*—§D°— 7D! + 11D + by = 0, wt 8. W841" D— Wy =0, es 8. if SF & mts, show thatx =, Gos me +c, sin me'+ cy cosh mt + cy sinh at,_ Lncan DirrEReNTiA, Eousnons FER INVERSE OPERATOR 1 (Definition. Fp, X is that function of x, not containing arbitrary constants which when operated upon by f(D) gives X. ie, or Thus 7o* satisfies the equation f (Dy = X and is, therefore, its particular integral. Obviously, f(D) and Wf (D) are inverse operators @) Let oi) Operating by D, DEx=Dy ie, x= 2 Integrating both sides w.r.t.x,y fxaz, no constant being added as (i) does not contain any constant. ‘Thus $x fxax. @) peak ne [xetax. Let pokey fii) Operating by D -a,(D—a). pig kW- avy a x= Pay, 40, & ay =X which is a Leibnite’s linear equation, LF. being e, its solution is yet'= [Xedx, no constant being added as (ji) doesn't contain any constant. Xayoe™ [xetds. [EEMGI RULES FOR FINDING THE PARTICULAR INTEGRAL which is symbolic form of (D? + k,D!! + kD? + atty a 5 son 29 Consider the equation > + k, at Fi ——___1_____. D'+hD'* +h DP? +..4h, Case I. When X = e** Since De® = ae Die = ae De = ret (DP + hy) +... + Re = (a" + Aya" +.. +h, ie, f (Die =f (aleOperating on both sides by 7, ok Dye = 7H Oe or et =, fare . dividing by f(a), 1 i TD) Hay’ Provided fa) +0 of) If fla) = 0, the above rule fails and we proceed further. Since a is a root of A.B. f(D) =D" + k,D™-" +... +k, =0. 2. D~—aisa factor of f(D). Suppose /(D) = (D - a) ¢ (D), where § (a) #0. Then 1 aia 14 Fi” * Da aD)” ~D-a' aa) (extn echt de fem ee [By $13.5 (2) = ya Daa ea Jer-etae ry $13.5 (3) 1 te, eet axl get = gat" Ie a5" ies ae” "Fe" (2) 5 f(D) =(D~ a)¥(D) +1. KD) + Pla)=0x9la. » Ma) Iff (a) = 0, then applying (2) again, we get 1 eax = x21 ess, provided fta) +0 (8) 1 t(D) ra) 1 & Solution. — PLL= [Put D=1) = + _*-©. wwe" 45146 12 [aati rn then oD + Dir se8 zat Solution, P= ——1___je®* y asin x}- 1 _fe¥ se" -e] (D+2XD-1? (D+2xD-1P Let us evaluate each of these terms separately. 1 ge i 1 (D+ 2xD-1F" —> [Resolving into partial fractions} = oe fF dy = dO [ad —itam andre +4 log cos ax) 08 ax a Changing i to ~i, we have -iax {i pigmeeret lef ican] 1 Fale e+ Stoe coax] ee {Log sa] teri 2 =F in oe 4h ty = Gz leg cosax. cos ax. (iii) Hence the C.S. is y=c, ¢08 ax +c, sin ax + (Vax sin ax + (1/a2) cos ax log cos ax.ms Hien Encnweerns Marvenatios ECEIStoeE Salve : 1, PY 6% yoy <6 te™ ig 2 (wn, 2008) i Sa Oar F897 og TU, By 4.55 = Bosh x. Also find y when y= 0, % = Lota =0, 2, ea @x ode a 8, Se tnts = kms int +0 4 Tio 2B Bea sine 5, F¥ gd : fs 5 SR++ By = dents (hopal, 20028) 6 (AAD + By ain tee (Madras, 2000) #y ty og oF eae costs i 2008 4, Spree ner asin te, TD, 200) 8 Siyae $y =a cote, ase, 9. (D®=8D9+ 7D —3)y =e%cosh x (Nagarjuna, 2008) 10. Spy deve (Wagpuir, 2009) dy 4h 1, DA Dy= 204164 cone 4 2 (Mumbai, 2008) 1 FY-0514 25y -eesinees WU, 2006) d’y. gay ‘ 13. (D241? y =xt42sinx coe 3x, (Madras, 2006) 14, Freer (Bhopal, 2008) Ai ay 9 ; 0 1 Dee Dts tye cos Le. rajasthan, 2008) 16, FF «2D say ae cn (wtU, 2010) 1%, DP 4D + By =e* sin x + xe, (Reipur, 2005 ; Anna, 2002 S) dy dty aa PY 495 mates cos Be 19, ©Y oy sees cosh x, ane? vgs aa” 2, } 20 (D+ DEA Dy mate eainte, CRT. 2008). FAG) whinge, W.2U,20108) 22, (DP 2D41y=xeisx. (Rajasthan, 2006) 28, (Pr Dy ersiner@exde | | | Pa Seesmic (S.V.1.U,, 2009) 25; (D? + a%y'= tan as. EEE Wo OTHER METHODS OF FINDING P.I. I Method of variation of parameters. This method is quite general and applies to equations of the form "+ py’ +qy=X ~) : yoX UX oe where p, q, and X are functions of x. It gives PI. = — yy Pet: +y: 2 Sa (2) where y, and y, are the solutions of y" + py’ + ay =0 (3) ond W. | >| is called the Wronskian" of ¥4,¥o- 1 Proof. Let the C.F. of (1) bey = ey, + ey Replacing ¢,, , (regarded as parameters) by unknown functions u(x) and v(x), let the P.I. be y= Uy, + Uy, afd) Differentiating (4) w.r.t. x, we get y= uy’, + vy’p + UY, + V'Vy ‘*Named after the Polish mathematician and philosopher Hoene Wronsky (1778-1853).muy,’ + 0y9 on assuming that u’y, + v’y) = 0 Differentiate (4) and substitute in (1). Then noting that y, and y,, satisfy (3), we obtain uly! + vy =X A) Solving (6) and (7), we get nx ne 2X , . w= a ae where W=y,y4’ 929 Integrating u=—- pak dz,v= p= dz. Substituting these in (4), we get (2). Solution. Given equation in symbolic form is (D? + 4)y = tan 2x. @) To find CF. ts AE, is D? + 2 D=s2 ThusCF.is — y=c, cos 2x +c, sin 2x. (ii) To find P.I. Here y, = cos 2x, yp = sin 2x and X= tan 2x on y2|_| cos2x sin 2x Wel -| Seas 20s 2x| 77 Thus, Pi amy fx deny, [Xe Ww Ww =~ cos 2x eee ee a ae 1 ~ Peas 2x f(ser 2x eos Baldr +E sin 2x fsin 2x dx =~} eos del log (see 2x + tan 2x)—sin Bet — 4 sin Bx eos Be ~ } eos 2x log(soe 2x + tan 25) 1 Hence the C.8. is y =e, cos 2x + ¢, sin 2x cos 2x log (see 2x + tan 2x). Solution. Given equation is D®—1 = 2/1 +e") AE. is D?-1=0,D=+1, Herey,=e%y,=e7 and X: ‘Thus ae ff L-aler—er* aga ee) = ef |e - fe ( ree a eae) 7 San e [-e* + log (e* + 1)] ~ e* log (1 + e*) =~ 1+ o* log (0 + 1)—e* log (e+ 1) ne" + eye — 1 + €* log (e* + 1) —e* log (e+ 1), a]-e* log(1 +e) HeneeCS.isSolution Given equation is (26D + 9 =e? AE. is D?-6D+9=Oie(D-3"=0 + CF. =(c, +e" Here y, =e%,y,=xe" and X= ex? [% de] _| fe é xe i | [ae + 3xe™ Solution. Given equation in symbolic form is (D? — 2D + 1) y = e* log x (i) To find C.F. Its AE. is(D-1)*=0, « D=1,1 Thus CF. is y = (¢, + o2)e" (ii) To find PL. Herey, =e%, y= xe" and X= ef logs w=[% %2[[% | ae Howl let Atxet ‘Thus PL=-y, Pam ae v4 pada PRE NOB ce 5 got FE IRE dy = 08 fo log d+ 0" flog x dx é ene (Smee fi Sas] cee (eines fixes) 2 x =-e |< loge (Su r}e @(xlogx—x) $e Blogs —3) Hence CS isy =(e, +eg)et+ 1 s%e'(2logx—3), IL. Method of undetermined coefficients ‘To find the P.L. of f(D) y = X, we assume a trial solution containing unknown constants which are deter- mined by substitution in the given equation. The trial solution to be assumed in each ease, depends on the form of X. Thus when (i) X= 2%, trial solution = ae. (i). X =8 sin 2x, trial solution = a, sin 2x + a, cos 2e 2x5, trial solution = a,x9 + a,x? + a,x +0, However when X= tan x or sec x, this method fails, since the number of terms obtained by differentiating X = tan x or see x is infinite. ‘The above method holds so long as no term in the trial solution appears in the C.F. Ifany term of the trial solution appears in the C.F., we multiply this trial solution by the lowest positive integral power of x which is large enough so that none of the terms which are then present, appear in the C.F.Solution. Here CF. =e* (c, cos (3 x+0, sin 3 x) ‘Assume PLL as ax? tax ta, +a" Dy = 20x +a, —a,e* and D¥y = 2a, + a,e* Substituting these in the given equation, we get dar? + (a, + Aa,) x + (2a, + 2d + dag) + Baje* = 2x” + Se* ‘Equating corresponding coefficients on both sides, we get da, = 2, 4a, + day = 0, 2a, + 2a, + 4a, = 0, 3a, = a, =0,a,= 1. Thus PL ahaa pave CS. is y =e*(c, cos \8x +0, sin \3z) + dat -taxwes, Solution. Here C.P.=c, cos x +c, sin x We would normally assume a trial solution as a, cos.x +a, sin x. ‘However, since these terms appear in the C-F., we multiply by x and assume the trial P.I. as y=x(@, cosx +a, sin x) “ Dy = (a, + 4,x) cos x + (ay ~ a,x) sin x and Dy = (2a, ~ a,x) cos x — (2a, + ayx) sin x ‘Substituting these in the given equation, we get 2a, cos x ~ 2a, sin x = sin x ‘Equating corresponding coefficients, 2a,=0, -2a,=1 sothata,=0,a,=— le C8. isy se, cos +69 sinx— > xsinz, Solution. Its A.E.isD?-1=0, . D=+1. ‘Thus CF. = ¢,e" + ¢¢ Assume P.I. asy = e¥ (c, cos 2x + ¢, sin 2x) ~e™(c, cos Bx + ¢, sin Bx) & = &% ((Be, + 2c,) cos 2x + (Bc, — 2e,) sin 2x} — e% (Be, + 3e,) cos Bx + (Ze, — Be,) sin Ax) and 4 8 ((Go, + 12¢q) 008 2x + (Be, ~ 12c,) sin Bx) ~e% (12c, ~ 6¢,) c08 3x ~ (Se, + 12c,) sin Bx} Substituting these in the given equation, we get e* ((de, + 12¢q) cos Be + (Ae, ~ 12c,) sin 2x] — e% {(12e, — 64) eos 3x — (Ge, + 12c,) sin Bx =e cos 2x — e sin Bx Equating corresponding coefficients, Ae, + 12¢y = 1, dey — 12¢ = 0; 12c, — Gey = 0, Ge, + 12ey=—1 whence e, = 1/40, ¢, = 9/40,¢,=-V/15,c,=— 1/80 Thus PLL = 2 e* (cos 2x +3 sin 2x) + 2 e% (2.008 8x + sim 8x) L Hence C.8. is y = ¢,e" + eye* Br 2(2.c08 e+ sin Sx) + Je (eos Bx + 8 sin 2),Ea Hicker Exaneerina MareManics: Eee REE ‘Solve by the miethod of-variation of parameters : P dl L a boy wewecen 2 GF ty=secx (Bhopal, 2007) _ 2 a Fw yatanx. (PTW, 2005 ; Raipur, 2004) Ps SY ey -zaine (SVT... 2007; LN-TU, dy dy 6 Fate tree WT. W2U, 20108; UPL, 2005) Ty" Bl 4 By =e tana (VT, 2010) 8. PTY, 2003) ag Sah od, 9) aaa 1, 2004) Solve by the method of undetermined coefficients = 10. @P~8D+2)y =x +e. (VP, 2003 8) 1. a += Zeon x, (EU, 2000.8) 2 12, aot deyedey cing win 2008) as. Ws 010) 14, (D2 = 2D +3) y= 29 + cos x. 15. D® = 2D) y = e* sinx. (WU, 2000) [EEE EQUATIONS REDUCIBLE TO LINEAR EQUATIONS WITH CONSTANT COEFFICIENTS Now we shall study two such forms of linear differential equations with variable coefficients which can be reduced to linear differential equations with constant coefficients by suitable substitutions. I. Cauchy's homogeneous linear equation*. An equation of the form dy : Ye hy= 4 hg gE + Bg =X a) where X is a function of x, is called Cauchy's homogeneous linear equation. Such equations can be reduced to linear differential equations with constant coefficients, by putting ‘ d or t=logx. Then ifD = © log. a dt dl Sng a tes fy_ a (ia)_1% ~- Ld ld’y dt_ 1 fd'y a) © dx \x dt) ~ 3? de “3a zat de Plae ad ie, 3 DW) —-1)y. Similarly, x* 3 = DID -1)(D-2)y and soon, After making these substitutions in (1), there results a linear equation with constant coefficients, which can be solved as before. 2 Example 13,30. Solve x° oy zx % y's loge! (V.2U,, 2020) Solution. This is a Cauchy's homogeneous linear. *See footnote p. 144.Putz=el,ie,t=logx, so thatz % = Dy, 2? £Y = Dw —1yy whereD = 4 hey ee a ‘Then the given equation becomes [D(D-1)-D +l y=t or (D-1? y=t wai) which is a linear equation with constant coefficients. Its AE. is (D - 1)? = 0 whence D = 1, 1. 1 CR.= (+ a PL (cy + cof et an DoF Hence the solution of (i) is y = (c, + cafe! + £ +2 or, putting ¢ = log x and y= (c, +6, logx) x + logx + 2 as the required solution of (i). t=(1-Dy*t=(1+2D + 3D? 4...) t=t+ 2. =x, we get Solution, Puts ¢! ie, ¢= logs so that x dle = Dy, x¢ 2 =D(D-Dy ‘Then the given equation becomes 1 -1)+3D +1] 2D+1)y= (D(D-1)+3D +1, ace or (D?+2D+1)y Its AE. is D?4+2D+1=0 or (D+1%=0ie,D= OF =(, tex) et = fe +epog2) 2 ott Ew I. eg ge MF pene Be a-ge Dal MO Tr a ae ae TRG which is Leibnitz’s linear equation having LF. = ot its solution is Solution, Putting x = e' ie. t = log x, the given equation becomes (D(D-1)+D+ily=tsint ie (D?+ Dy=tsint WO Its AE. is D?+1=Oie. D=2i. cy ¢08 ¢ +c, sint and 1_, “D+ 2D)ca Hi Ec MAE) =Un ote saat ag tour bet $12)" =LP. ote By .) #1. or got E(e+8) =LP. oe (t+ 5) av=ue. «5 (5-3) =n ote ( peg) — 1),“Lunean Dirrenerne Eauaroxs L495 | y =e, (2x 1) +09 (Be — 1? + : (ae -1P + ; (2x ~ 1) log (2-1) -2. which is the required solution, PS yO gas 8 PTD any Vay x8 (Sv. du, du 6. "The radial displacement in a rotating dice at a distance r from the axis is given by 2 2H+r4 ys a8 = 0, where k is a constant, Solve the equation under the conditions «= 0when r=0,u=0whenr =a. Solve : ay | gg a dy | ds 6 2 ee + 2y =logx. (Bhopal, 2009) z eorree cate +y=x+logx (Bhopal, 2008) 8. xy" 4 ay" + y = Boos? log x), (W.T.U, 2011) (S.V.T.U, 2006); P.1:U,, 2008) PTW, 2008) 13, ey JoeB + dy aloes, (UP.2.U, 2008) ye hy at sa ae (Bhopa?, 2008) de de 7 3 ody. ore 4 ‘13, (2x +3) er ei ane (V. EU, 2007 ; Kerala, 2005 ; Anna, 2002 8) dy a ta = 1 22s oye 2)? epee vers trees (Wagpur, 2009) oy dx 1 (taste + (i+) ® ¥ ymin tog cae ai) (PTW, 2006; V.-7.1r, 2004) oS & TU, 2006; V-T., ya By Be ses a (Murbai, 2006) 16. (Gx 9 2 Sp eB (Bes 2) Ze [EEBEG (1) LINEAR DEPENDENCE OF SOLUTIONS Consider the initial value problem consisting of the homogeneous linear equation x + py tqy=0 with variable coefficients p (x) and g (x) and two initial conditions y (xg) = ky. ¥' (xp) = hy Let its general solution be y = ¢,, + €29'2 which is made up of two linearly dependent solutions y, and y,.* If plx) and q(x) are continuous functions on some open interval I and x, is any fixed point on I, then the above initial value problem has a unique solution y(x) on the interval I. + Asin §2.12, 9,9, are an interval /, if and only if there exist numbers 2, 2, not both zero such that Ay, + 2a). = Ino such numbers other than zero exist, then y,,, are said to be linearly independent.ca Hictie ENCES MATEARES) (2) Theorem. If p(x) and q(x) are continuous on an open interval I, then the solutions y, and y, of (1) are linearly dependent in 1 if'and only if the Wronskian' WY y ¥,) for some x, on J. If there is an x= 2, in at which W (y, y,)#0, then y,, y, are linearly independent on I. Proof: Ify,.9 are linearly dependent solutions of (1) then there exist two constants ¢,,¢, not both zero, such that eat eye 0 44) Differentiating wir.t.x, cy," + ey,/=0 (8) Eliminating cy, ¢, from (4) and (6), we get W Oye¥9) 2 a im Conversely, suppose W (¥,, 4) = 0 for some x =z, on J and show that y,,,79 are linearly dependent. ear Gea) + carat) Consider the equation ; 7 (6) e1¥i lo) + eay9(xo} | nl) alo) Yi) ¥2l%9) Hence the system has a solution in which ¢,, ¢, are not both zero. Now introduee the funetion (x) = cyy,(x) + cy) ‘Then y(x) is a solution of (1) on Z. By (6), this solution satisfies the initial conditions y (xq) = 0 and y'(x,) = 0. Also since PG) and q(x) are continuous on J, this solution must be unique, But y = 0 is obviously ancther solution of (1) satisfying the given initial conditions. Hence ¥ =y ie., cy) + €,¥= 0 in I. Now since ey, ¢, are not both zero, it implies that y, andy, are linearly dependent on f. which, on eliminating ¢,, cy, give W (,.¥2) Example 19,97. Show that the two functions sin 2x, cos 2x are independent solutions of y7 4 4y = 0. Solution, Substituting y, = sin 2« (ory, = cos 2x) in the given equation we find that y,,, are its solutions. sindx eos Also 10) = 240 Ww) | seosax —2ein 2x * for any value of x, Hence the solutions y,, 9 are linearly independent, eeeees Solve : 1, Show that e~*, xe"* are independent solutions of y” + 2y’ +y = 0 in any.interval. ‘2. Show that e* cos x, e* sin x are independent solutions of the equation xy”~ 2y' = 0. 8. Ify,, 9, be two solutions ofy” + p(x) y’ + g(x) =0, show that the Wronskian can be expressed as WOy,7.)= ae SIMULTANEOUS LINEAR EQUATIONS WITH CONSTANT COEFFICIENTS Quite often we come across linear differential equations in which there are two or more dependent variables and a single independent variable. Such equations are known as simultaneous linear equations. Here we shall deal with systems of linear equations with constant coefficients only. Such a system of equations is solved by eliminating all but one of the dependent variables and then solving the resulting equations as before. Each of the dependent variables is obtained in a similar manner. Example 13.38. Solve the simultaneous equations : ae dy at +5 —2y=t, ae +2x+y=0 being given x = y= 0 when t = 0. (S.V.T.U., 2009 ; Kurukshetra, 2005) + See footnote on p. 486.