Differential Equation Solver Simulator For Runge-Kutta Methods
Differential Equation Solver Simulator For Runge-Kutta Methods
Differential Equation Solver Simulator For Runge-Kutta Methods
1, 2016 RESEARCH
DOI: 10.17482/uujfe.70981
Metin HATUN *
Fahri VATANSEVER *
Öz: Mühendislik ve fen bilimlerine ait problemlerinin birçoğu diferansiyel denklemlerle matematiksel
olarak modellenmekte, dolayısıyla çözümleri önemli yer tutmaktadır. Diferansiyel denklemlerin analitik
veya sayısal çözümleri için çok sayıda yöntemler geliştirilmiştir. Teknolojinin gelişmesiyle orantılı olarak
sayısal çözüm yöntemlerinden yoğun bir şekilde faydalanılmaktadır. Özellikle gerçek zamanlı
uygulamalarda en az işlemle, en kısa sürede, en yüksek hassasiyetle, en doğru çözüme ulaşmak başlıca
hedeflerdendir. Gerçekleştirilen çalışmada, diferansiyel denklemlerin sayısal çözümü için Runge-Kutta
tabanlı 48 yöntemi içeren bir simülatör geliştirilmiştir. Kullanıcı dostu ve eğitim amaçlı da
kullanılabilecek simülatörde tanımlanan diferansiyel denklemin, belirtilen başlangıç koşulu altında,
verilen adım boyutu veya nokta sayısına göre, istenen aralıktaki çözümü seçilen yöntemle elde
edilebilmektedir. Çözümler kullanıcıya hem sayısal (adım değerleri, hesaplama süresi vb.) hem de
grafiksel olarak sunulabilmekte; bunun yanında yöntemlerle/çözümlerle ilgili konu anlatımları da
verilebilmektedir. Ayrıca simülatörle karşılaştırmalı çözümler (performans analizleri) de
gerçekleştirilebilmekte. Böylece simülatör ile kullanıcılar farklı yöntemlerle diferansiyel denklemlerin
sayısal çözümlerini gerçekleştirebilmekte; öğrenciler bu alandaki yöntemleri konu anlatımı destekli görsel
olarak öğrenip adım adım uygulayabilmekte; tasarımcılar da karşılaştırmalı analizlerle sistemleri için en
uygun yöntemi kolay, etkin ve doğru bir şekilde seçebilmektedirler.
*
Uludag University, Faculty of Engineering, Electrical-Electronics Engineering Department, 16059 Bursa/Turkey
Corresponding author: Fahri Vatansever, (fahriv@uludag.edu.tr)
145
Hatun M.,Vatansever F: Differential Equation Solver Simulator for Runge-Kutta Methods
1. INTRODUCTION
2. RUNGE-KUTTA METHODS
A general first order ordinary differential equation (ODE) system under the initial
condition 𝑦0 = 𝑦(𝑥0 ) can be given as
To obtain a second order Runge-Kutta method, Taylor series expansion can be used as
ℎ 2 ′′
𝑦(𝑥 + ℎ) = 𝑦(𝑥) + ℎ𝑦 ′ (𝑥) + 2
𝑦 (𝑥) + 𝒪(ℎ3 ) (2)
𝑦 ′ (𝑥) ≅ 𝑓(𝑥, 𝑦)
𝜕𝑓 𝜕𝑓 𝜕𝑦 } (3)
𝑦 ′′ (𝑥) = + . = 𝑓𝑥 + 𝑓𝑦 . 𝑦 ′ = 𝑓𝑥 (𝑥, 𝑦) + 𝑓𝑦 (𝑥, 𝑦). 𝑓(𝑥, 𝑦)
𝜕𝑥 𝜕𝑦 𝜕𝑥
ℎ2
𝑦(𝑥 + ℎ) = 𝑦(𝑥) + ℎ𝑓(𝑥, 𝑦) + 2
[𝑓𝑥 (𝑥, 𝑦) + 𝑓𝑦 (𝑥, 𝑦). 𝑓(𝑥, 𝑦)] + 𝒪(ℎ3 )
ℎ ℎ
} (4)
𝑦(𝑥 + ℎ) = 𝑦(𝑥) + 2 𝑓(𝑥, 𝑦) + 2 [𝑓(𝑥, 𝑦) + ℎ𝑓𝑥 (𝑥, 𝑦) + ℎ𝑓𝑦 (𝑥, 𝑦)𝑓(𝑥, 𝑦)] + 𝒪(ℎ3 )
146
Uludağ University Journal of The Faculty of Engineering, Vol. 21, No.1, 2016
𝑓(𝑥 + ℎ, 𝑦 + 𝑓(𝑥, 𝑦)) = 𝑓(𝑥, 𝑦) + ℎ𝑓𝑥 (𝑥, 𝑦) + ℎ𝑓𝑦 (𝑥, 𝑦)𝑓(𝑥, 𝑦) + 𝒪(ℎ2 ) (6)
𝑘1 = 𝑓(𝑥𝑖 , 𝑦𝑖 ) ℎ
} ⇒ 𝑦𝑖+1 = 𝑦𝑖 + 2 (𝑘1 + 𝑘2 ) (8)
𝑘2 = 𝑓(𝑥𝑖 + ℎ, 𝑦𝑖 + ℎ𝑘1 )
and also known as first order Runge-Kutta method (Vatansever, 2006). Many different methods
have been developed with similar steps. The used methods with their iteration equations in the
designed simulator are given in Table 1-5.
147
Hatun M.,Vatansever F: Differential Equation Solver Simulator for Runge-Kutta Methods
k1 f ( xi , yi ) k1 f ( xi , yi )
1 1 1 1
k 2 f xi h , yi k1h k 2 f xi h , yi k1h
3 3 2 2
2 2 3 3
k3 f xi h , yi k 2 h k3 f xi h , yi k 2 h
3 3 4 4
yi 1 yi
1
k1 3k3 h yi 1 yi 2k1 3k 2 4k3 h
1
4 9
3rd order contraharmonic mean method (Ababneh
3rd order harmonic mean method (Wazwaz, 1990)
and Rozita, 2009)
k1 f ( xi , yi ) k1 f ( xi , yi )
2 2 2 2
k 2 f xi h , yi k1h k 2 f xi h , yi k1h
3 3 3 3
2 2 4 2 2
k3 f xi h , yi k1h k 2 h k3 f xi h , yi k 2 h
3 3 3 3 3
1 2k1k 2 2k 2 k 3 1 k12 k 22 k 22 k32
h
yi 1 yi h yi 1 yi
2 k1 k 2 k 2 k3 2 k1 k 2
k 2 k3
3rd order modified contraharmonic mean weights method (Ababneh and 3rd order arithmetic mean Runge-Kutta method
Rozita, 2009) (Wazwaz, 1994; Ahmad and Yaacob, 2013)
k1 f ( xi , yi ) k1 f ( xi , yi )
2 2 2 2
k 2 f xi h , yi k1h k 2 f xi h , yi k1h
3 3 3 3
4 4 2 1
k3 f xi (3 2 )h , yi (3 2 )k 2 h k3 f xi h , yi k1h k 2 h
21 21 3 3
1 k12 k 22 3 k 22 k32 1 k1 k 2 k 2 k3
yi 1 yi h yi 1 yi h
4 k1 k 2 4 k 2 k3
2 2 2
New 3rd order arithmetic mean Runge-Kutta method (Ahmad and Yaacob, Composite arithmetic–harmonic mean Runge–Kutta
2013) method (Ahmad and Yaacob, 2005)
k1 f ( xi , yi ) k1 f ( xi , yi )
2 2 2 2
k 2 f xi h , yi k1h 3 3
k 2 f xi h , yi k1h
3 3
5 5
1 1 k1 k 2 4 4 k1 k 2
k3 f xi h , yi h k3 f xi h , yi ( )h
3 3 2 5 5 2
43 2 3(4 3 2 ) 3(2 2 ) 1 2k1k 2 2k 2 k 3
yi 1 yi k1 k2 yi 1 yi h
2 k1 k 2 k 2 k3
k3 h
4 4 2
148
Uludağ University Journal of The Faculty of Engineering, Vol. 21, No.1, 2016
k1 f ( xi , yi )
k1 f ( xi , yi )
1 1
1 1 k 2 f xi h , yi k1h
k 2 f xi h , yi k1h 2 2
3 3 3 1
2 3 k3 f xi h , yi (k1 5k 2 )h
k3 f xi h , yi k1h k 2 h 8 8
3 4 1
k 4 f xi h , yi (k1 k 2 k3 )h k 4 f xi h , yi (5k1 7k 2 18k3 )h
20
yi 1 yi
1
k1 3k 2 3k3 k 4 h 1 2k1k 2 2k 2 k 3 2k 3 k 4
8 yi 1 yi h
3 k1 k 2 k 2 k3 k3 k 4
4th order Runge-Kutta method with minimum error bound 4th order contraharmonic mean Runge-Kutta method (Ababneh
(Ralston, 1962) and Rozita, 2009)
k1 f ( xi , yi )
1 1
k 2 f xi h , yi k1h
k1 f ( xi , yi ) 2 2
k 2 f ( xi 0.4h , yi 0.4k1h )
5 1 3
k 3 f xi 0.45573725 h , yi (0.29697761k1 0.15875964 k 2 )h k3 f xi h , yi k1h k 2 h
8 8 8
k 4 f xi h , yi (0.21810040 k1 3.05096516 k 2 3.83286476 k 3 )h
yi 1 yi (0.17476028k1 0.55148066k 2 1.20553560k 3
1 3 3
k 4 f xi h , yi k1h k 2 h k3 h
0.17118478k 4 )h 4 4 2
1 k k 2 k 2 k3 k3 k 4
2 2 2 2 2 2
yi 1 yi 1 h
3 k1 k 2 k 2 k3 k3 k 4
4th order heronian mean mehtod (Murugesan et. al., 2001) RK4: Butcher-Johnston (Butcher and Johnston, 1993)
k1 f ( xi , yi )
k1 f ( xi , yi )
1 1
k 2 f xi h , yi k1h 3 3
2 2 k 2 f xi h , yi k1h
4 4
23 1 25
k 3 f xi h , yi k1h k 2 h 1 1 1
48 48 48 k3 f xi h , yi k1h k 2 h
2 3 6
1 47 289
k 4 f xi h , yi k1h k2h k3 h 1 2
24 600 300 k 4 f xi h , yi k1h k 2 h 2k3 h
3 3
1
yi 1 yi k1 2k 2 2k 3 k 4 k1k 2 k 2 k 3 k 3 k 4 h
yi 1 yi k1 4k3 k 4 h
1
9
6
4th order centroidal mean method (Murugesan et. al., 2001) RK4: England (England, 1969)
k1 f ( xi , yi ) k1 f ( xi , yi )
1 1
k 2 f ( xi h , yi k1h ) 1 1
2 2 k 2 f xi h , yi k1h
13 1 11
2 2
k 3 f ( xi h , yi k1h k2 h ) 1 1
24 24 24 k3 f xi h , yi k 2 h
1 25 73 4 4
k 4 f ( xi h , yi k1h k2h
k 4 f xi h , yi k 2 h 2k3 h
k3 h )
12 132 66
2 k k1k 2 k 2 k 2 k 2 k3 k32 k32 k3 k 4 k 42
2 2 2
yi 1 yi 1 h yi 1 yi
1
k1 4k3 k 4 h
9 k1 k 2 k 2 k3 k3 k 4
6
RK4: Gill (Gill, 1951)
k1 f ( xi , yi )
1 1
k 2 f xi h , yi k1h
2 2
1 1
k3 f xi h , yi k2 h
1 1
k1h 1
2 2 2 2
k 4 f xi h , yi k3 h
1 1
k 2 h 1
2 2
1 1 1 1 1 1
yi 1 yi k1h 1 k 2 h 1 k3 h k 4 h
6 3 2 3 2 6
149
Hatun M.,Vatansever F: Differential Equation Solver Simulator for Runge-Kutta Methods
150
Uludağ University Journal of The Faculty of Engineering, Vol. 21, No.1, 2016
k1 f ( xi , yi )
4 4
k 2 f xi h , yi k1h
11 11
2 1
k3 f xi h , yi (9k1 11k 2 ) h
5 50
1 1
k 4 f xi h , yi (11k1 15k3 )h
2 4
k5 f xi
6 6
10
h , yi
1
600
(81 9 6 )k1 (255 55 6 )k3 (24 14 6 )k 4 h
k 6 f xi
6 6
10
h, yi
1
600
(81 9 6 )k1 (255 55 6 )k3 (24 14 6 )k 4
h
yi 1 yi
1
36
4k1 (16 6 )k5 (16 6 )k 6 h
151
Hatun M.,Vatansever F: Differential Equation Solver Simulator for Runge-Kutta Methods
k1 f ( xi , yi ) k1 f ( xi , yi )
k 2 f xi h , yi k1h k 2 f xi h , yi k1h
1 1 1
k3 f xi h , yi (k1 k 2 )h k3 f xi h , yi (3k1 k 2 )h
2 2 8
1 1 1
k 4 f xi h , yi (14k1 5k 2 3k3 )h k 4 f xi h , yi (k1 k 2 4k3 )h
4 64 2
1 1 1 1
k5 f xi h , yi (12k1 12k 2 8k3 64k 4 )h k5 f xi h , yi (4k1 5k 2 20k3 3k 4 )h
2 96 4 64
3 1 3 1
k 6 f xi h , yi (9k 2 5k3 16k 4 36k5 )h k6 f xi h , yi (12k1 9k 2 12k3 7k 4 32k5 )h
4 64 4 64
yi 1 yi
1
7k1 7k3 32k 4 12k5 32k6 h yi 1 yi
1
7k1 12k3 7k 4 32k5 32k6 h
90 90
RK5: Luther-3 (Luther, 1966)
k1 f ( x i , y i )
k 2 f x i h , y i k 1 h
k 3 f x i 12 h , y i 18 (3k1 k 2 h
k 4 f x i h , y i 12 (k1 k 2 4k 3 )h
k 5 f xi 5 5
h , yi 1
(25 7 5 )k (5 5 5 )k (20 4 5 )k 2 5k h
f x (3 5 )k (1 5 )k (4 4 5 )k 2k 4 5k h
10 100 1 2 3 4
k6 i
5 5
10
h , yi 1
20 1 2 3 4 5
y i 1 y i 1
12
k1 k 4 5k 5 5k 6 h
RK5: Butcher-3 (Butcher, 1994, 1969) RK5: Butcher-4 (Butcher, 1994, 1969)
k1 f ( xi , yi ) k1 f ( xi , yi )
1 1 1 1
k 2 f xi h , yi k1h k 2 f xi h , yi k1h
5 5 5 5
2 2 2 2
k3 f xi h , yi k 2 h k3 f xi h , yi k 2 h
5 5 5 5
3 75 9 117 1 3 5
k 4 f xi h , yi k1h k 2 h k3 h k 4 f xi h , yi k1h k3 h
4 64 4 64 2 16 16
37 7 3 4 1 5
k5 f xi h , yi k1h k 2 h k3 h k 4 h k5 f xi h , yi k1h k3 h 2k 4 h
36 3 4 9 4 4
19 25 2 1 1 2 1
yi 1 yi k1 k3 k 4 k5 h yi 1 yi k1 k 4 k5 h
144 48 9 8 6 3 6
RK5: Butcher-5 (Butcher, 1994, 1969)
k1 f ( xi , yi )
1 1
k 2 f xi h , yi k1h
5 5
2 2
k3 f xi h , yi k 2 h
5 5
3 161 19 287
k 4 f xi h , yi k1h k 2 h k3 h
4 192 12 192
27 19 291 36
k5 f xi h , yi k1h k 2 h k3 h k 4 h
28 7 196 49
7 475 2 7
yi 1 yi k1 k3 k 4 k5 h
48 1008 7 72
RK5: Butcher-6 (Butcher, 2010)
152
Uludağ University Journal of The Faculty of Engineering, Vol. 21, No.1, 2016
k1 f ( xi , yi )
1 1
k 2 f xi h , yi k1h
10 10
1 11 41
k3 f xi h , yi k1h k 2 h
5 260 260
3 3 63 9
k 4 f xi h , yi k1h k2h k3 h
10 170 340 68
3 13035 75447 9009 1287
k5 f xi h , yi k1h k2h k3 h k 4 h
4 544 1088 136 64
165709 733747 638689 32053 5320
k 6 f xi h , yi k1h k2 h k3 h k4h k5 h
918 1326 1122 162 11583
5 250 32 1
yi 1 yi k1 k4 k5 k 6 h
54 567 81 14
153
Hatun M.,Vatansever F: Differential Equation Solver Simulator for Runge-Kutta Methods
k1 f ( xi , yi )
1 1
k 2 f xi h , yi k1h
2 2
2 2 4
k 3 f xi h , yi k1h k 2 h
3 9 9
1 7 2 1
k 4 f xi h , yi k1h k 2 h k 3 h
3 36 9 12
5 35 55 35 15
k 5 f xi h , yi k1h k 2 h k 3 h k 4 h
6 144 36 48 8
1 1 11 1 1 1
k 6 f xi h , yi k1h k 2 h k 3 h k 4 h k 5 h
6 360 36 8 2 10
41 23 43 118 32 80
k 7 f xi h , yi k1h k 2 h k3h k4h k 5 h k 6 h
260 13 156 39 195 39
13 11 11 4 4 13
yi 1 yi k1 k3 k4 k5 k6 k 7 h
200 40 40 25 25 200
RK6: Butcher-4 (Butcher, 1964)
k1 f ( xi , y i )
k 2 f xi h, y
5 5
k h 5 5
f x k h
10 i 10 1
5 5 5 2 5
k3 h, y k h 5
f x k h
i 10 i 10 1 10 2
15 7 5 1 5
k4 h, y
5 5
k h k h 15 7 5
f x k h
i 10 i 20 1 4 2 10 3
5 5 5 5 15 7 5
k5 h, y k h k h 1
f x k h
i 10 i 60 1 6 3 60 4
5 5 5 3 5
k6 h, y
5 5
k h k h k h 95 5 1
f x h , y k h
i 10 i 60 1 12 3 6 4 10 5
55 25 5 25 7 5 5 5
k7 i i
1
6
k h 1 k h (5 2 5 )k h
12 3 12 4 5 2
k6 h
y i k1 5k 5 5k 6 k 7 h
1
y i 1
12
RK6: Luther (Luther, 1967)
k1 f ( xi , yi )
k2 f xi h , yi k1h
1 1
k3 f xi h , yi (3k1 k2 )h
2 8
2 1
k4 f xi h , yi (8k1 2k2 8k3 )h
3 27
k5 f xi
7 21
14
h , yi
1
392
3 (3 21 7)k1 8(7 21 )k2 48(7 21 )k3 3(21 21 )k4 h
k6 f xi
7 21
14
h , yi
1
1960
5 (231 51 21 )k1 40(7 21 )k2 320 21 k3 3(21 121 21 )k4 392(6 21 )k5 h
k7 f xi h , yi
1
180
15 (22 7 21 )k1 120k2 40(7 21 5)k3 63(3 21 2)k4 14(49 9 21 )k5 70(7 21 )k6 h
yi 1 yi
1
9k1 64k3 49k5 49k6 9k7 h
180
The main screenshot and descriptions of the simulator designed by using MATLAB
(Mathworks, 2007) are given in Figure 1. The ordinary differential equations, initial condition,
solutions range and step size or number of points selectively, are entered to the simulator. The
solution of the differential equation is implemented according to the selected method, and the
solution values/steps are also listed. In addition, the solution of the differential equation in the
specified range is drawn graphically. The solution range is also determined by the user. By
using menu options of the simulator (see Table 6), the results can be saved or printed; single or
comparative solutions can be created; program settings (drawing shape of data, graphical data
representation formats, etc.) can be made (see Table 7) and help topics can be accessed. In
addition, the simulator calculate the solutions for the entered ODE using three methods selected
154
Uludağ University Journal of The Faculty of Engineering, Vol. 21, No.1, 2016
by the user and presents the obtained results comparatively (solution values,
computation/elapsed times etc.) to the user on the separate windows both as numerical and
graphical. As an example, the solution steps of the simulator by using a second order Runge-
Kutta method are summarized with a flowchart given in Figure 2. Similar steps are carried out
with the other methods.
Figure 1:
Main screen of designed simulator
Pan
New
Save
Zoom In
3D rotate
Zoom Out
Data cursor
155
Hatun M.,Vatansever F: Differential Equation Solver Simulator for Runge-Kutta Methods
R
Start
f(x,y) , x0 , y0 , a , b , h or n
E C
x=x0 , y=y0
Stop Figure 3:
RC circuit
Figure 2:
Flowchart for Euler's method
In the first simulation, the following differential equation belong to a RC circuit given in
Figure 3 is used.
1 1
𝑦 ′ (𝑡) + 𝑅𝐶 𝑦(𝑡) = 𝑅𝐶 𝐸 (10)
The screenshot of the solution of the differential equation of the RC circuit, for 𝑅 = 1𝑀Ω,
𝐶 = 1𝜇𝐹, 𝐸 = 12𝑉 values, under the initial condition 𝑦(0) = 0 within solution range 0 − 10𝑠
with 3rd order Runge-Kutta method-1 method is given in Figure 4. The comparative error
results for the first simulation are given in Table 8.
Figure 4:
Screenshot of single solution simulation
156
Uludağ University Journal of The Faculty of Engineering, Vol. 21, No.1, 2016
The screenshot for the comparative solution of the ODE given above is given in Figure 5.
157
Hatun M.,Vatansever F: Differential Equation Solver Simulator for Runge-Kutta Methods
Figure 5:
Screenshot of comparative solution simulation
4. CONCLUSIONS
In the study, a general purpose simulator which can make numerical solution of ordinary
differential equations with Runge-Kutta-based methods and be used in different technical fields
(basic and applied sciences, engineering sciences, etc.) has been developed. In the simulator
which can be used also for educational purposes; numerical solution of ODE defined/entered by
the user is obtained by selected method using the specified initial condition and within the
specified range, and the obtained results can be shown as both numerical and graphical. In the
simulator which contains 48 different methods, the subject explanations are also included. In
addition, comparative numerical solutions of ODE can also be studied with the methods chosen
by the user. Thus, the users can easily show the effectiveness and efficiency of the methods by
using the simulator which presents comparative numerical solutions of ODEs, and the use of the
simulator can facilitate the choice (for using, for learning etc.) of the method for the students or
those who are going to use the methods in their applications. As a future work, it is planned to
expand the simulator by adding the other methods presented for numerical solution of ODEs,
and to present it for online use by preparing in the form of a web page.
158
Uludağ University Journal of The Faculty of Engineering, Vol. 21, No.1, 2016
REFERENCES
1. Ababneh, O.Y. and Rozita, R. (2009) New third order Runge Kutta based on
contraharmonic mean for stiff problems, Applied Mathematical Sciences, 3(8), 365-376.
2. Abraham, O. and Bolarin, G. (2011) On error estimation in Runge-Kutta methods,
Leonardo Journal of Sciences, 10(18), 1-10.
3. Ahmad, R.R. and Yaacob, N. (2005) Third-order composite Runge–Kutta method for stiff
problems, International Journal of Computer Mathematics, 82(10), 1221-1226. doi:
10.1080/00207160512331331039
4. Ahmad, R.R. and Yaacob, N. (2013) Arithmetic-mean Runge-Kutta method and method of
lines for solving mildly stiff differential equations, Menemui Matematik (Discovering
Mathematics), 35(2), 21-29.
5. Butcher, J.C. (1964) On Runge-Kutta processes of higher order, Journal of the Australian
Mathematical Society, 4(2), 179-197.
6. Butcher, J.C. (1969) The effective order of Runge-Kutta methods, Lecture Notes in
Mathematics, 109, 133-139.
7. Butcher, J.C. and Johnston, P.B. (1993) Estimating local truncation errors for Runge-Kutta
methods, Journal of Computational and Applied Mathematics, 45(1-2), 203-212.
doi:10.1016/0377-0427(93)90275-G
8. Butcher, J.C. (1994) Initial value problems: numerical methods and mathematics,
Computers and Mathematics with Applications, 28(10-12), 1-16. doi:10.1016/0898-
1221(94)00182-0
9. Butcher, J.C. (1996) A history of Runge-Kutta methods, Applied Numerical Mathematics,
20(3), 247-260. doi:10.1016/0168-9274(95)00108-5
10. Butcher, J.C. (2000) Numerical methods for ordinary differential equations in the 20th
century, Journal of Computational and Applied Mathematics, 125(1-2), 1-29.
doi:10.1016/S0377-0427(00)00455-6
11. Butcher, J.C. (2009) On fifth and sixth order explicit Runge-Kutta methods: order
conditions and order barriers, Canadian Applied Mathematics Quarterly, 17(3), 433-445.
12. Butcher, J.C. (2010) Trees and numerical methods for ordinary differential equations,
Numerical Algorithms, 53(2-3), 153-170. 10.1007/s11075-009-9285-0
13. Chapra, S.C. and Canale, R.P. (2002) Numerical Methods for Engineers: with Software and
Programming Applications, 4th ed., McGraw-Hill, New York.
14. Derr, L., Outlaw, C., and Sarafyan, D. (1993) A new method for derivation of continuous
Runge-Kutta formulas, Computers and Mathematics with Applications, 26(3), 7-13.
doi:10.1016/0898-1221(93)90105-5
15. England, R. (1969) Error estimates for Runge-Kutta type solutions to systems of ordinary
differential equations, The Computer Journal, 12(2), 166-170.
16. Evans, D.J. (1989) New Runge-Kutta methods for initial value problems, Applied
Mathematics Letters, 2(1), 25-28. doi:10.1016/0893-9659(89)90109-2
17. Gill, S. (1951) A process for the step-by-step integration of differential equations in an
automatic digital computing machine, Mathematical Proceedings of the Cambridge
Philosophical Society, 47(1), 96-108.
159
Hatun M.,Vatansever F: Differential Equation Solver Simulator for Runge-Kutta Methods
160
Uludağ University Journal of The Faculty of Engineering, Vol. 21, No.1, 2016
39. Ralston, A. (1962) Runge-Kutta methods with minimum error bounds. Mathematics of
Computation, 16(80), 431-437. doi:http://dx.doi.org/10.1090/S0025-5718-1962-0150954-0
40. Ralston, A. and Rabinowitz, P. (1978) A First Course in Numerical Analysis, 2nd ed.,
McGraw-Hill, New-York.
41. Vatansever, F. (2006) İleri Programlama Uygulamaları, Seçkin Yayıncılık, Ankara.
42. Wazwaz, A.M. (1990) A modified third order Runge-Kutta method, Applied Mathematics
Letters, 3(3), 123-125. doi:10.1016/0893-9659(90)90154-4
43. Wazwaz, A.M. (1991) Modified numerical methods based on arithmetic and geometric
means, Applied Mathematics Letters, 4(3), 49-52. doi:10.1016/0893-9659(91)90034-S
44. Wazwaz, A.M. (1994) A comparison of modified Runge-Kutta formulas based on a variety
of means, International Journal of Computer Mathematics, 50(1-2), 105-112. doi:
10.1080/00207169408804245
161
162