Numerical analysis

Numerical analysis for engineer by Stephen Chapra

Newton Raphson- method ( solution of equation by iteration)

Newton Raphson- method:

Is a method of root location formula?

Xi= is initial guess at the root

f ’(Xi)= is the tangent of the curve at the

point [Xi , f (Xi) ]

′(
𝑓 ′ (𝑥𝑖 − 0)
𝑓 𝑥𝑖 ) = tan 𝜃 =
𝑥𝑖 − 𝑥𝑖+1
𝑓(𝑥𝑖 )
𝑥𝑖+1 − 𝑥𝑖 = − … … . (∗)
𝑓 ′ (𝑥𝑖 )
𝐸𝑥: 𝑏𝑒𝑔𝑖𝑛𝑖𝑛𝑔 𝑤𝑖𝑡ℎ 𝑎𝑛 𝑖𝑛𝑖𝑡𝑖𝑎𝑙 𝑒𝑠𝑡𝑖𝑚𝑎𝑡𝑒 𝑜𝑓 𝑥𝑖 = 5 , 𝑒𝑠𝑡𝑖𝑚𝑎𝑡𝑒 𝑡ℎ𝑒 𝑠𝑖𝑛𝑔𝑙𝑒
𝑟𝑜𝑜𝑡 𝑜𝑓 𝑡ℎ𝑒 𝑒𝑞𝑢𝑎𝑡𝑖𝑜𝑛
𝑒 𝑥 − 10 cos 𝑥 − 100 = 0
𝑠𝑜𝑙: 𝑓 (𝑥𝑖 ) = 𝑒 𝑥 − 10 𝑐𝑜𝑠 𝑥 − 100 … … . . (1)
𝑓 ′ (𝑥𝑖 ) = 𝑒 𝑥 + 10 sin 𝑥 … … … (2)

𝑖 𝑥𝑖 𝑓 (𝑥𝑖 ) 𝑓 ′ (𝑥𝑖 ) 𝑥𝑖+1 − 𝑥𝑖

1 5 45.5765 138.823 −0.3283
2 4.67169 7.28509 96.88 -0.0752
3 4.596 0.292 89.2 -0.00328
4 4.593 0.00052 88.882 -0.000006
5 4.593 −2 ∗ 10−8 88.881 2.25 ∗ 10−10
𝐸𝑥: 𝑢𝑠𝑒 𝑡ℎ𝑒 𝑛𝑒𝑤𝑡𝑜𝑛 𝑅𝑎𝑝ℎ𝑠𝑜𝑛 𝑚𝑒𝑡ℎ𝑜𝑑 𝑡𝑜 𝑒𝑠𝑡𝑖𝑚𝑎𝑡𝑒 𝑡ℎ𝑒 𝑟𝑜𝑜𝑡 𝑜𝑓 𝑒 −𝑥 − 𝑥
𝑒𝑚𝑝𝑙𝑜𝑦𝑖𝑛𝑔 𝑎𝑛 𝑖𝑛𝑖𝑡𝑖𝑎𝑙 𝑔𝑢𝑒𝑠𝑠 𝑜𝑓 𝑥𝑖 = 0
𝑠𝑜𝑙: 𝑓 (𝑥 ) = 𝑒 −𝑥 − 𝑥 ……..(1)

𝑓 ′(𝑥) = −𝑒 −𝑥 − 1…………..(2)
𝑓(𝑥𝑖 )
𝑥𝑖+1 − 𝑥𝑖 = − … … . (∗)
𝑓 ′ (𝑥𝑖 )

𝑖 𝑥𝑖 𝑓 (𝑥𝑖 ) 𝑓′(𝑥𝑖 ) 𝑥𝑖+1 − 𝑥𝑖

1 0 1 -2 0.5
2 0.5 0.10653 -2.64872 -0.0402
 Taylor series
𝑃𝑘 (𝑥 ) = 𝑎0 + 𝑎1 𝑥 + 𝑎2 𝑥 2 + 𝑎3 𝑥 3 + ⋯ + 𝑎𝑘 𝑥 𝑘 (𝑝𝑜𝑙𝑦𝑛𝑜𝑚𝑖𝑎𝑙)
𝑤ℎ𝑒𝑟𝑒; 𝑎0 , 𝑎1 , 𝑎2 , … , 𝑎𝑘 : 𝑎𝑟𝑒 𝑐𝑜𝑒𝑓𝑓𝑖𝑐𝑖𝑒𝑛𝑡
𝑙𝑒𝑡 𝑘 = 3 → 𝑃3 (𝑥 ) = 𝑎0 + 𝑎1 𝑥 + 𝑎2 𝑥 2 + 𝑎3 𝑥 3
𝑤ℎ𝑒𝑟𝑒 𝑎0 , 𝑎1 , 𝑎2 , 𝑎3 , … , 𝑎𝑘 : 𝑎𝑟𝑒 𝑐𝑜𝑒𝑓𝑓𝑖𝑐𝑖𝑒𝑛𝑡 𝑤ℎ𝑖𝑐ℎ 𝑎𝑟𝑒 𝑒𝑣𝑎𝑙𝑢𝑎𝑡𝑒𝑑
𝑓𝑟𝑜𝑚 𝑝𝑜𝑖𝑛𝑡 [0, 𝑓(0)]
𝑃3 (0) = 𝑎0 = 𝑓 (0)
𝑃′3 (𝑥 ) = 𝑎1 + 2𝑎2 𝑥 + 3𝑎3 𝑥 2
𝑃′3 (0) = 𝑎1 = 𝑓 ′ (0)
𝑓 ′ (0) = 𝑎1
𝑃′′3 (𝑥 ) = 2𝑎2 + 6𝑎3 𝑥

′′ (
𝒇′′ (𝟎)
𝑃′′3 (0) = 2𝑎2 = 𝑓 0 ) → 𝒂𝟐 =
𝟐!
𝑃′′′3 (𝑥 ) = 6𝑎3

′′′ (
𝒇′′′ (𝟎)
𝑃′′′3 (0) = 6𝑎3 = 𝑓 0) → 𝒂𝟑 =
𝟑!

′(
𝒇′′ (𝟎) 𝟐 𝒇′′′ (𝟎) 𝟑
𝑷𝟑 (𝒙) = 𝒇(𝟎) + 𝒇 𝟎)𝒙 + 𝒙 + 𝒙
𝟐! 𝟑!

𝑤ℎ𝑒𝑛 𝑥 → ∞ 𝑡ℎ𝑒 𝑝𝑜𝑙𝑦𝑛𝑜𝑚𝑖𝑎𝑙 𝑐𝑎𝑙𝑙𝑒𝑑 𝑚𝑎𝑐𝑙𝑢𝑟𝑖𝑎𝑛 𝑠𝑒𝑟𝑖𝑒𝑠

𝐸𝑥: 𝑓𝑖𝑛𝑑 𝑇𝑎𝑦𝑙𝑜𝑟 𝑝𝑜𝑙𝑦𝑛𝑜𝑚𝑖𝑎𝑙 𝑝𝑛 (𝑥 )𝑔𝑒𝑛𝑒𝑟𝑎𝑡𝑒𝑑 𝑏𝑦 𝑓 (𝑥 ) = 𝑒 𝑥 𝑎𝑡 𝑥 = 0
𝑠𝑜𝑙: 𝑇𝑎𝑦𝑙𝑜𝑟 𝑝𝑜𝑙𝑦𝑛𝑜𝑚𝑖𝑎𝑙

𝑓 ′ (0) 𝑓 ′′ (0) 2 𝑓 ′′′ (0) 3

𝑓(𝑥 ) = 𝑓(0) + 𝑥+ 𝑥 + 𝑥 +⋯
1! 2! 3!
𝑓(𝑥 ) = 𝑒 𝑥 , 𝑓 (0) = 1
𝑓 ′ (𝑥 ) = 𝑒 𝑥 , 𝑓 ′ (0) = 1
𝑓 ′′ (𝑥 ) = 𝑒 𝑥 , 𝑓 ′′ (0) = 1
𝑓 ′′′ (𝑥 ) = 𝑒 𝑥 , 𝑓 ′′′ (0) = 1

𝑥
𝑥2 𝑥3
𝑒 =1+𝑥+ + +⋯
2! 3!
If instead of approximating the value of f(x) near zero, we are concerned with value
of x near some other point a.

𝑃𝑛 (𝑥 ) = 𝑎0 + 𝑎1 (𝑥 − 𝑎) + 𝑎2 (𝑥 − 𝑎)2 + ⋯ + 𝑎𝑛 (𝑥−𝑎)𝑛
∴ 𝑡ℎ𝑒 𝑇𝑎𝑦𝑙𝑜𝑟 𝑠𝑒𝑟𝑖𝑒𝑠 𝑔𝑒𝑛𝑒𝑟𝑎𝑡𝑒𝑑 𝑏𝑦 𝑓 𝑎𝑡 𝑥 = 𝑎

′(
𝑓 ′′ (𝑎) 2
𝑓 ′′′ (𝑎)
𝑓(𝑥 ) = 𝑓(𝑎) + 𝑓 𝑎) ∗ (𝑥 − 𝑎) + (𝑥 − 𝑎) + (𝑥 − 𝑎)3 +
2! 3!
𝑓 𝑛 (𝑎)
…+ (𝑥 − 𝑎)𝑛
𝑛!
𝑤ℎ𝑒𝑛 𝑥 → ∞ 𝑡ℎ𝑒 𝑝𝑜𝑙𝑦𝑛𝑜𝑚𝑖𝑎𝑙 𝑐𝑎𝑙𝑙𝑒𝑑 𝑡𝑎𝑦𝑙𝑜𝑟 𝑠𝑒𝑟𝑖𝑒𝑠.
𝐸𝑥: 𝑢𝑠𝑒 𝑧𝑒𝑟𝑜 𝑡ℎ𝑟𝑜𝑢𝑔ℎ 𝑓𝑜𝑢𝑟𝑡ℎ 𝑜𝑟𝑑𝑒𝑟 . 𝑇𝑎𝑦𝑙𝑜𝑟 𝑠𝑒𝑟𝑖𝑒𝑠 𝑒𝑥𝑝𝑎𝑛𝑠𝑖𝑜𝑛 𝑡𝑜 𝑎𝑝𝑝𝑟𝑜𝑥𝑖𝑚𝑎𝑡𝑒 𝑡ℎ𝑒

𝑓𝑢𝑛𝑐𝑡𝑖𝑜𝑛:

𝑓(𝑥 ) = −0.1 𝑥 4 − 0.15 𝑥 3 − 0.5 𝑥 2 − 0.25 𝑥 + 1.2

𝑓𝑟𝑜𝑚 𝑥𝑖 = 0 𝑤𝑖𝑡ℎ ℎ = 1
𝑡ℎ𝑎𝑡 𝑝𝑟𝑒𝑑𝑖𝑐𝑡 𝑡ℎ𝑒 𝑓𝑢𝑛𝑐𝑡𝑖𝑜𝑛 𝑣𝑎𝑙𝑢𝑒 𝑎𝑡 𝑥 = 1
𝑠𝑜𝑙: 𝑓𝑟𝑜𝑚 𝑡𝑎𝑦𝑙𝑜𝑟 𝑠𝑒𝑟𝑖𝑒𝑠 ∶
𝑓 ′ (𝑎) 𝑓 ′′ (𝑎) 2
𝑓 ′′′ (𝑎)
𝑓(𝑥 ) = 𝑓(𝑎) + (𝑥 − 𝑎) + (𝑥 − 𝑎) + (𝑥 − 𝑎)3 +
1! 2! 3!
𝑓 ′′′′ (𝑎)
(𝑥 − 𝑎)4 + ⋯
4!
𝑓(𝑎) = 𝑓(0) = 1.2
𝑓 ′ (𝑎) = 𝑓 ′ (0) = −0.25
𝑓 ′′ (𝑎) = 𝑓 ′′ (0) = −1
𝑓 ′′′ (𝑎) = 𝑓 ′′′ (0) = −0.9
𝑓 ′′′′ (𝑎) = 𝑓 ′′′′ (0) = −2.4
𝑓 ′′′′′ (𝑎) = 𝑓 ′′′′′ (0) = 0
𝑓𝑟𝑜𝑚 𝑡𝑎𝑦𝑙𝑒𝑟 𝑠𝑒𝑟𝑖𝑒𝑠
0.25 1 0.9 2.4
𝑓(𝑥 ) = 1.2 − (1 − 0) − (1 − 0)2 − (1 − 0)3 − (1 − 0)4
1! 2! 3! 4!
𝑓(𝑥 ) = 0.2
𝐸𝑥: 𝑢𝑠𝑒 𝑇𝑎𝑦𝑙𝑜𝑟 𝑠𝑒𝑟𝑖𝑒𝑠 𝑡𝑜 𝑎𝑝𝑝𝑟𝑜𝑥𝑖𝑚𝑎𝑡𝑒 √17 𝑏𝑒𝑔𝑖𝑛𝑖𝑔 𝑓𝑟𝑜𝑚 √16

𝑠𝑜𝑙: 𝑒𝑥𝑎𝑐𝑡 √17 = 4.1231056

𝑇𝑎𝑦𝑙𝑜𝑟 𝑠𝑒𝑟𝑖𝑒𝑠
𝑓 ′ (𝑎) 𝑓 ′′ (𝑎) 2
𝑓 ′′′ (𝑎)
𝑓(𝑥 ) = 𝑓(𝑎) + (𝑥 − 𝑎) + (𝑥 − 𝑎) + (𝑥 − 𝑎)3 + ⋯
1! 2! 3!
𝑓(𝑥 ) = √𝑥 → 𝑓(0) = 0

𝑓(𝑎) = √16 = 4
Interpolation :
If a table of values of function f(x) is given it is often necessary to obtain values of
f(x) for value of x between the x-values appearing in the table.

1. Newton’s Divided – Difference Interpolation

a) Linear interpolation:
The simplest form of interpolation is to connect two data point
with a straight line.

𝑙𝑒𝑡 𝑓1 (𝑥 ) = 𝑎0 + 𝑎1 𝑥 … . (∗)
𝑤ℎ𝑒𝑟𝑒 𝑓1 (𝑥 ) ∶ 𝑙𝑖𝑛𝑒𝑎𝑟 𝑖𝑛𝑡𝑒𝑟𝑝𝑜𝑙𝑎𝑡𝑖𝑜𝑛 𝑓𝑜𝑟𝑚𝑢𝑙𝑎 .

𝑤ℎ𝑒𝑛 𝑥 = 𝑥0 → 𝑓1 (𝑥 ) = 𝑓(𝑥0 )
𝑓 (𝑥0 ) = 𝑎0 + 𝑎1 𝑥0 … … … (1)
𝑤ℎ𝑒𝑛 𝑥 = 𝑥1 → 𝑓1 (𝑥 ) = 𝑓(𝑥1 )
𝑓 (𝑥1 ) = 𝑎0 + 𝑎1 𝑥1 … … … (2)
𝑓𝑟𝑜𝑚 𝑒𝑞 (1)& (2)
𝑓 (𝑥1 ) − 𝑓(𝑥0)
𝑎1 =
𝑥1 − 𝑥0
𝑓(𝑥1 ) − 𝑓(𝑥0 )
𝑎0 = 𝑓(𝑥0 ) −
𝑥1 − 𝑥0
𝑓𝑟𝑜𝑚 𝑒𝑞 (∗)
𝑓(𝑥1 ) − 𝑓(𝑥0 )
𝑓(𝑥 ) = 𝑓(𝑥0 ) + (𝑥 − 𝑥0 )
𝑥1 − 𝑥0
𝐸𝑥: 𝑒𝑠𝑡𝑖𝑚𝑎𝑡𝑒 𝑡ℎ𝑒 𝑙𝑛2 𝑢𝑠𝑖𝑛𝑔 𝑙𝑖𝑛𝑒𝑎𝑟 𝑖𝑛𝑡𝑒𝑟𝑝𝑜𝑙𝑎𝑡𝑖𝑜𝑛 . 𝑢𝑠𝑖𝑛𝑔 ∶
1) ln 1 = 0 , ln 6 = 1.79176
2) ln 1 = 0 , ln 4 = 1.386294
𝑓(𝑥1 ) − 𝑓(𝑥0 )
𝑠𝑜𝑙: 𝑓(𝑥 ) = 𝑓(𝑥0 ) + (𝑥 − 𝑥0 )
𝑥1 − 𝑥0
𝑢𝑠𝑖𝑛𝑔 𝑥 = 1 → ln 1 = 0
𝑥 = 6 → ln 6 = 1.79176
1.79176 − 0
ln 2 = 0 + ∗ (2 − 1)
6−1
ln 2 = 0.3583
𝑢𝑠𝑖𝑛𝑔 𝑤ℎ𝑒𝑛 𝑥 = 1 , ln 1 = 0
𝑥 = 4 , ln 4 = 1.386294
1.386294 − 0
ln 2 = 0 + (2 − 1)
4−1
ln 2 = 0.46209
From calculator ln 2 = 0.69314
b) Quadratic interpolation :
If three data point are available this can be accomplished with a
second order polynomial
𝑓2 (𝑥 ) = 𝑎0 + 𝑎1 𝑥+𝑎2 𝑥 2 … … … (∗∗)
𝑤ℎ𝑒𝑛 𝑥 = 𝑥0 , 𝑓2 (𝑥 ) = 𝑓(𝑥0 )
𝑥 = 𝑥1 , 𝑓2 (𝑥 ) = 𝑓(𝑥1 )
𝑥 = 𝑥2 , 𝑓2 (𝑥 ) = 𝑓(𝑥2 )
𝑓(𝑥0 ) = 𝑎0 + 𝑎1 𝑥0 +𝑎2 𝑥0 2 … … … (1)
𝑓(𝑥1 ) = 𝑎0 + 𝑎1 𝑥1 +𝑎2 𝑥1 2 … … … (2)
𝑓(𝑥2 ) = 𝑎0 + 𝑎1 𝑥2 +𝑎2 𝑥2 2 … … … (3)

𝑓2 (𝑥 ) = 𝑏0 + 𝑏1 (𝑥 − 𝑥0 ) + 𝑏2 (𝑥 − 𝑥1 )(𝑥 − 𝑥0 )
𝑤ℎ𝑒𝑟𝑒; 𝑏0 = 𝑓(𝑥0 )

𝑓(𝑥1 ) − 𝑓(𝑥0 )
𝑏1 =
𝑥1 − 𝑥0

𝒇(𝒙𝟐 ) − 𝒇(𝒙𝟏 ) 𝒇(𝒙𝟏 ) − 𝒇(𝒙𝟎 )

−
𝒙𝟐 − 𝒙𝟏 𝒙𝟏 − 𝒙𝟎
𝒃𝟐 =
𝒙𝟐 − 𝒙𝟎
𝐸𝑥: 𝐸𝑠𝑡𝑖𝑚𝑎𝑡𝑒 𝑡ℎ𝑒 𝑛𝑎𝑡𝑢𝑟𝑎𝑙 𝑙𝑜𝑔𝑎𝑟𝑖𝑡ℎ𝑚 𝑜𝑓 2 𝑢𝑠𝑖𝑛𝑔 𝑞𝑢𝑎𝑑𝑟𝑎𝑡𝑖𝑐
𝑖𝑛𝑡𝑒𝑟𝑝𝑜𝑙𝑎𝑡𝑖𝑜𝑛
𝑥0 = 1 → 𝑙𝑛1 = 0
𝑥1 = 4 → 𝑙𝑛4 = 1.3863
𝑥2 = 6 → 𝑙𝑛6 = 1.7917
𝑠𝑜𝑙: 𝑏0 = 𝑓(𝑥0 ) = ln 1 = 0
𝑓(𝑥1 ) − 𝑓(𝑥0 ) 1.3863 − 0
𝑏1 = → 𝑏1 =
𝑥1 − 𝑥0 4−1
𝑏1 = 0.46209

𝑓 (𝑥2 ) − 𝑓(𝑥1 ) 𝑓(𝑥1 ) − 𝑓(𝑥0 )

−
𝑥2 − 𝑥1 𝑥1 − 𝑥0
𝑏2 =
𝑥2 − 𝑥0

1.7917 − 1.3863 1.3863 − 0

− 4−1
𝑏2 = 6 − 4
6−1

𝑏2 = −0.05187116
𝑓2 (𝑥 ) = 𝑏0 + 𝑏1 (𝑥 − 𝑥0 )+𝑏2 (𝑥 − 𝑥1 )(𝑥 − 𝑥0 )
= 0 + 0.46209(𝑥 − 1) − 0.05187116(𝑥 − 4)(𝑥 − 1)
x=2
𝑙𝑛 2 ≅ 𝑓2 (𝑥 ) = 0.5658446
𝑓1 (𝑥 ) = 0.46209
𝑙𝑛 2 = 0.69314
 Numerical Method for Solving First Order D.E
Step by step:
We stand from 𝑦0 = 𝑦(𝑥0 ) and proceed stepwise:

Step by step method

Euler Method İmproved Euler method Runge-kutta method

Euler Method
From Taylor series :-

′(
ℎ2 ′′
𝑦(𝑥 + ℎ) = 𝑦(𝑥 ) + ℎ 𝑦 𝑥 ) + 𝑦 (𝑥 ) + ⋯
2!
If we take only two term:

𝑦(𝑥 + ℎ) = 𝑦(𝑥 ) + ℎ 𝑦 ′ (𝑥 )
𝑑𝑦
= 𝑓(𝑥, 𝑦)
𝑑𝑥
Where h is very small
𝑦(𝑥 + ℎ) ≌ 𝑦(𝑥 ) + ℎ𝑓

Where h ≌ step
The step of computing stant
𝑥 = 𝑥0 , 𝑦 = 𝑦0
1. 𝑦1 ≌ 𝑦0 + ℎ 𝑓(𝑥0 , 𝑦0 )
𝑓(𝑥0 , 𝑦0 ) = 𝑠𝑡𝑟𝑒𝑝 𝑜𝑓 𝑐𝑢𝑟𝑣𝑒 𝑎𝑡 𝑥0 , 𝑦0
2. 𝑦2 ≌ 𝑦1 + ℎ 𝑓(𝑥1 , 𝑦1 )
𝑓(𝑥0 , 𝑦0 ) = 𝑠𝑙𝑜𝑝𝑒 𝑜𝑓 𝑡ℎ𝑒 𝑐𝑢𝑟𝑣𝑒 𝑎𝑡 𝑥1 , 𝑦1
..
..
..
𝑦𝑛+1 ≌ 𝑦𝑛 + ℎ 𝑓(𝑥𝑛 , 𝑦𝑛 )

Ex: Apply Euler method to initial value problem

𝑦 ′ − 𝑥 − 𝑦 = 0, 𝑦(0) = 0
Choosing, h=0.2, and computing 𝑦1 → 𝑦5
Sol
𝑑𝑦
= 𝑥 + 𝑦 = 𝑓(𝑥, 𝑦)
𝑑𝑥
𝑥0 = 0, 𝑦0 =0
𝑦𝑛+1 ≌ 𝑦𝑛 + ℎ 𝑓(𝑥𝑛 , 𝑦𝑛 )
𝑦𝑛+1 = 𝑦𝑛 + ℎ (𝑥𝑛 + 𝑦𝑛 )
𝑦𝑛+1 = 𝑦𝑛 + 0.2 (𝑥𝑛 + 𝑦𝑛 )

n 𝑥𝑛 𝑦𝑛 0.2 (𝑥𝑛 + 𝑦𝑛 ) 𝑦𝑛+1

0 0 0 0 0
1 0.2 0 0.04 0.04
2 0.4 0.04 0.088 0.128
3 0.6 0.128 0.1456 0.274
4 0.8 0.274 0.215 0.489
5 1 0.489
 Improved Euler method
𝑑𝑦
= 𝑓(𝑥, 𝑦)
𝑑𝑥

𝒉
In the internal from xn to xn+ , we approximate the solution y by the straight
𝟐
line. Through (xn , yn), with slope f(xn,yn) and then we continue along the
straight line with slope f (xn+1, y*n+1), untile x reach to xn+1

y*n+1 = yn + h f(xn , yn)

yn+1 = yn + 𝒉𝟐 {f (xn , yn) + f(xn+1, y*n+1)}

Ex: Apply the improved Euler method to the initial value problem:
y’ = x + y
y(0) = 0 choosing h = 0.2 y1 → y5
sol:
h = 0.2
y*n+1 = yn + h f(xn , yn)
y*n+1 = yn +0.2 (xn+ yn)
improved Euler
yn+1 = yn + 𝒉𝟐 {f (xn , yn) + f(xn+1, y*n+1)}
𝟎.𝟐
yn+1 = yn + { xn + yn+ xn+1 + yn + 0.2 (xn + yn)}
𝟐

yn+1 = 1.22 yn + 0.12 xn + 0.1 xn+1

n xn yn 0.12 xn 0.1 xn+1 1.22 yn yn+1

0 0 0 0 0.02 0 0.02
1 0.2 0.02 0.024 0.04 0.0244 0.0884
2 0.4 0.0884 0.048 0.06 0.1078 0.2158
3 0.6 0.2158 0.072 0.08 0.2633 0.4153
4 0.8 0.4153 0.096 0.1 0.5067 0.7027
5 1 0.7027
 Fourth order Runge-kutta Method:
𝟏
yn+1= yn + { (k1 + 2 k2 + 3 k3 + k4)} h
𝟔

where : k1= f (xn , yn)

𝟏 𝟏
k2= f (xn + h , yn + h k1)
𝟐 𝟐
𝟏 𝟏
k3= f (xn + h , yn + h k2)
𝟐 𝟐
k4= f (xn + h , yn + h k3)

Ex: Apply the Runge-kutta method to the initial value problem

𝒅𝒚
= 𝒙 + 𝒚, y(0)=0, choosing h= 0.2 and computing five step.
𝒅𝒙

Sol:
𝒅𝒚
= 𝒙 + 𝒚 = 𝒇(𝒙, 𝒚)
𝒅𝒙

k1= f (xn , yn)

k1= xn + yn
𝟏 𝟏
k2= f (xn + h , yn + h k1)
𝟐 𝟐
𝟎.𝟐 𝟎.𝟐
k2= xn + + yn + (xn + yn)
𝟐 𝟐

k2= xn + 𝟎. 𝟏 + yn + 𝟎. 𝟏 k1
𝟏 𝟏
k3= f (xn + h , yn + h k2)
𝟐 𝟐

k3= xn + 𝟎. 𝟏 + yn + 𝟎. 𝟏 k2)
k4= f (xn + h , yn + h k3)
k4= xn + 0.2 + yn + 0.2 k3)
𝟏
yn+1= yn + { (k1 + 2 k2 + 3 k3 + k4)} h
𝟔

n xn yn k1 k2 k3 k4 yn+1
0 0 0 0 0.1 0.11 0.222 0.0214
1 0.2 0.0214 0.2214 0.3435 0.3557 0.4926 0.09182
2 0.4 0.09182 0.4918 0.641 0.6559 0.823 0.2221
3 0.6 0.2221 0.8221 1.0043 1.0225 1.2266 0.4255
4 0.8 0.4255 1.2255 1.448 1.47 1.7196 0.718
5 1 0.718