skf2133-chapter5n

Upload: samandondon

Post on 01-Jun-2018

217 views

Category:

Documents


0 download

TRANSCRIPT

  • 8/9/2019 skf2133-chapter5n

    1/26

    Page 5 - 1

    Chemical Engineering Numerical Method

      copyright PCS- FKKKSA, UTM

    Numerical Methods forNumerical Methods for

    Chemical EngineersChemical EngineersChapter 5: Curve FittingChapter 5: Curve Fitting

    SaharudinSaharudin HaronHaron

  • 8/9/2019 skf2133-chapter5n

    2/26

    Page 5 - 2

    Chemical Engineering Numerical Method

      copyright PCS- FKKKSA, UTM

    Curve FittingCurve Fitting

    Data is often given as a discrete values along a continuum and

     problems may occur when one need to estimate points between thediscrete values.

    This chapter describes techniques to fit curves to such data to obtain

    intermediate estimates.

    There are 2 general approaches :

    a) Least-squares regression

     b) Interpolation

  • 8/9/2019 skf2133-chapter5n

    3/26

  • 8/9/2019 skf2133-chapter5n

    4/26

    Page 5 - 4

    Chemical Engineering Numerical Method

      copyright PCS- FKKKSA, UTM

    Linear Regression Fitting a straight line that represents the general trend of the data.

     y = a o

    + a1 x + e

    where

    e = error or residual n = number of points

    ( )

     x xn

     x x

    n

     x

     x x

     x x

     of meanarithmetic 

    yof meanarithmeticy yy

    intercepta a-ya

    slopea -n

    y-yn a

    i

    i

    o1o

    12i

    2i

    iiii1

    =∑

    =

    =∑=

    ==

    =∑ ∑

    ∑ ∑ ∑=

  • 8/9/2019 skf2133-chapter5n

    5/26

    Page 5 - 5

    Chemical Engineering Numerical Method

      copyright PCS- FKKKSA, UTM

    Example 5.1:

    Fit a straight line from the Table 1 below.

    Table 1

    xi 1 2 3 4 5 6 7

    yi 0.5 2.5 2.0 4.0 3.5 6.0 5.5

    Solution: n = 7 Σ   xi yi = 119.5 Σ   xi2 = 140   Σ   yi = 24   Σ   xi = 28

     x = 28/7 = 4 a1 = 0.8392857

     y = 24/7 = 3.42857   a0 = 0.07142857

    Therefore, the least-squares fit is

     y = 0.01742857 + 0.8392857x

    Linear Regression

     x

     y

  • 8/9/2019 skf2133-chapter5n

    6/26

    Page 5 - 6

    Chemical Engineering Numerical Method

      copyright PCS- FKKKSA, UTM

    Quantification of Error of Linear Regression Standard deviation, Sy

    ( )∑   −=−=

    2

    t

    t

    y

     s

    1

    S

     s

     y y

    n

    i

    - the measurement of spread of data around the mean

    - the sum of the squares of the residuals between the

    data points and the mean

    Variance, Sy2 → (standard deviation)2

    1

    S s t

    2

    y−

    =n

    Regression error, Sr 

    ( )∑ ∑= =

    ==n

    1

    n

    1i

    2

    i1oi

    2

    ir  xa-a-yesi

    - the sum of the squares of the residuals

     between the measured and calculated y

  • 8/9/2019 skf2133-chapter5n

    7/26

    Page 5 - 7

    Chemical Engineering Numerical Method

      copyright PCS- FKKKSA, UTM

    Quantification of Error of Linear Regression Standard error of the estimate

    2

    S

     sr 

    y/ −= n x- the measurement of spread of data around

    the regression line

    Coefficient of determination

    t

    r t2S

    SS r    −= - shows the efficiency of the estimation of spread of the data points using regression line compare to the mean value

    tcoefficienncorrelatio- r r

    or 

    2=

    i.e. r 2 = 0.923 - indicate that 92.3 % of the original uncertainty has been

    explained by the linear model

  • 8/9/2019 skf2133-chapter5n

    8/26

  • 8/9/2019 skf2133-chapter5n

    9/26

    Page 5 - 9

    Chemical Engineering Numerical Method

      copyright PCS- FKKKSA, UTM

    Quantification of Error of Linear Regression

    The standard deviation is:

    sy

    = √[(22.7143 / (7 – 1)] = 1.9457

    The standard error is:

    s y/x = √ [(2.9911 / (7-2)] = 0.7735

    Thus because Sy/x < Sy, the linear regression ,model has merit.

    r 2 = (St – Sr ) / St = 22.7143 – 2.9911/ (22.7143) = 0.868

    or 

    r = √r 2 = √(0.868)2 = 0.932

    The results indicate that 86.8% of original uncertainty has been

    explained by the linear model.

  • 8/9/2019 skf2133-chapter5n

    10/26

  • 8/9/2019 skf2133-chapter5n

    11/26

  • 8/9/2019 skf2133-chapter5n

    12/26

    Page 5 - 12

    Chemical Engineering Numerical Method

      copyright PCS- FKKKSA, UTM

    Interpolation• To estimate intermediate values between precise data points. The most

    common method used for this purpose is polynomial interpolation.

    • General formula for an nth-order polynomial

    • f(x) = a o + a1 x + a 2 x 2 + …….. + a n x

     n

    • Several methods of interpolating polynomials such as :

    •1st order (linear) connecting 2 points.•2nd order (quadratic or parabolic) connecting 3 points.

    •3rd order (cubic) connecting 4 points.

    • The techniques that will be used are:-•Newton’s Interpolating Polynomials

    •Lagrange Interpolating Polynomial

    •Splines Interpolation

  • 8/9/2019 skf2133-chapter5n

    13/26

  • 8/9/2019 skf2133-chapter5n

    14/26

    Page 5 - 14

    Chemical Engineering Numerical Method

      copyright PCS- FKKKSA, UTM

    Quadratic (or parabola) Interpolation

    Connecting 3 points with a curve.

    f 2( x) = b0 + b1( x- x0) + b2( x- x0)( x- x1)

    where :

     b0

    = f( x0)

    f( x1) - f( x0)

     x1 - x0

    f( x2) - f( x1) - f( x1) - f( x0)

      x2 - x1   x1 - x0

     x2 –  x0

     b1 = f [ x1, x0] =

     b2 = f [ x2, x1, x0] =

  • 8/9/2019 skf2133-chapter5n

    15/26

    Page 5 - 15

    Chemical Engineering Numerical Method

      copyright PCS- FKKKSA, UTM

    Cubic Interpolation (third order)

    Connecting 4 points with a curve.

    f 3( x) = f( x

    0) + ( x- x

    0) f[ x

    1, x

    o] + ( x- x

    0)( x- x

    1) f [ x

    2, x

    1, x

    o]

    + ( x- x0)( x- x1)( x- x2) f [ x3, x2, x1, xo]

    where :

    f( x1) - f( x0) x1 - x0

    f [ x2, x1] - f [ x1, x0]

     x2 - x0

    f [ x3 , x2, x1] - f [ x2, x1, x0]

     x3

    - x0

    f [ x1, x0] =

    f [ x2

    , x1

    , x0

    ] =

    f [ x3, x2, x1, x0] =

  • 8/9/2019 skf2133-chapter5n

    16/26

    Page 5 - 16

    Chemical Engineering Numerical Method

      copyright PCS- FKKKSA, UTM

    Newton’s Divided-Difference Interpolating

    Polynomials – assignment in class

    Table 1 shows the relationship between applied stress and the time tofracture for a stainless steel:

    Fracture time, h 5 10 15 20 25 30

    Applied stress,kg/mm2

    11.6 10.3 9.1 8.2 7.4 6.8

    Estimate the fracture time for an applied stress of 8.5 kg/mm

    2

    . Use Newtoninterpolation of order 3.

  • 8/9/2019 skf2133-chapter5n

    17/26

    Page 5 - 17

    Chemical Engineering Numerical Method

      copyright PCS- FKKKSA, UTM

    Lagrange Interpolating Polynomials Is a reformulation of the Newton polynomial that avoids the

    composition of divided differences.

    It can be represented as :

    n

     f n( x) = ∑ Li( x) f ( xi)

    i=0

    where  Li( x) = ∏

    =   −−n

    i j

     j   ji

     j

     x x

     x x

    0

  • 8/9/2019 skf2133-chapter5n

    18/26

    Page 5 - 18

    Chemical Engineering Numerical Method

      copyright PCS- FKKKSA, UTM

    Lagrange Interpolating Polynomials Example :

    If n = 1

    If n = 2

    )()()( 101

    00

    10

    11   x f 

     x x

     x x x f 

     x x

     x x x f 

    −−+

    −−=

    )())((

    ))((

    )())((

    ))(()(

    ))((

    ))(()(

    21202

    10

    12101

    200

    2010

    212

     x f  x x x x

     x x x x

     x f  x x x x

     x x x x x f 

     x x x x

     x x x x x f 

    −−

    −−+

    −−

    −−+

    −−

    −−=

  • 8/9/2019 skf2133-chapter5n

    19/26

    Page 5 - 19

    Chemical Engineering Numerical Method

      copyright PCS- FKKKSA, UTM

    Splines Interpolation

    3 cases:

    Linear Splines Quadratic Splines

    Cubic Splines

    Linear Splines

    - is a simplest method where a straight line is drawn to connect 2 points.

    - a group of ordered data points can be defined as a set of linear functions.

  • 8/9/2019 skf2133-chapter5n

    20/26

    Page 5 - 20

    Chemical Engineering Numerical Method

      copyright PCS- FKKKSA, UTM

    A set of linear functions can be shown as below :-

     f ( x) = f ( x0) + m0( x- x0)   x0 ≤  x ≤  x1

     f ( x) = f ( x1) + m1( x- x1)   x1 ≤  x ≤  x2

    :

    :

     f ( x) = f ( xn-1) + mn-1( x- xn-1)  xn-1 ≤  x ≤  xn

    where

    mi = f ( xi+1) - f ( xi)

     xi+1 - xi

    Splines Interpolation

  • 8/9/2019 skf2133-chapter5n

    21/26

    Page 5 - 21

    Chemical Engineering Numerical Method

      copyright PCS- FKKKSA, UTM

    Splines Interpolation Example : Fit the data below with first order Splines. Evaluate the

    function at x = 5

    Solution : For the interval from x = 4.5 to x = 7.0

    m = 2.5 - 1.0

    7.0 - 4.5

    So, f (5) = (1.0) + (0.6)(5.0-4.5)

    = 1.3

     x f(x)3.0 2.5

    4.5 1.0

    7.0 2.5

    9.0 0.5

    = 0.6

  • 8/9/2019 skf2133-chapter5n

    22/26

    Page 5 - 22

    Chemical Engineering Numerical Method

      copyright PCS- FKKKSA, UTM

    Splines Interpolation

    Quadratic Splines

    the objective in quadratic splines is to derive a second order polynomial

    for interval between data points.

    The polynomial for each interval can be represented generally as :

     f i(x) = ai x 2 + bi x + ci - general equation for Quadratic Splines

    For n+1 data points, ( i = 0, 1, 2,….n) there are n intervals and

    consequently, 3n unknowns constants (the a’s, b’s and c’s) to evaluate.

  • 8/9/2019 skf2133-chapter5n

    23/26

    Page 5 - 23

    Chemical Engineering Numerical Method

      copyright PCS- FKKKSA, UTM

    Splines Interpolation

    The function values must be EQUAL at the interior knots. This conditioncan be represented as :

    ai-1 x2

    i-1 + bi-1 xi-1 + ci-1 = f ( xi-1)

    ai x2 i-1 + bi xi-1 + ci = f ( xi-1)for i = 2 to n.

    The first and last functions must pass through the end points.

    a1 x20 + b1 x0 + c1 = f ( x0)

    an x2n + bn xn + cn = f ( xn)

    Conditions that are required to evaluate the unknowns

  • 8/9/2019 skf2133-chapter5n

    24/26

  • 8/9/2019 skf2133-chapter5n

    25/26

    Page 5 - 25

    Chemical Engineering Numerical Method

      copyright PCS- FKKKSA, UTM

    Splines Interpolation

    Cubic Splines

    - the objective is to derive a third order polynomial for each interval. f i (x) = ai x

     3 + bi x 2 + ci x + d i

    For a quadratic splines, 4n conditions are required to evaluate the

    unknowns.These are :

    The function values must be equal at the interior knots (2n-2 conditions)

    The first and last functions must pass through the end points (2 conditions)

    The first derivatives at the interior knots must be equal (n-1 conditions) The second derivatives at the interior knots must be equal (n-1 conditions)

    The second derivatives at the end knots are zero (2 conditions)

  • 8/9/2019 skf2133-chapter5n

    26/26

    Page 5 - 26

    Chemical Engineering Numerical Method

      copyright PCS- FKKKSA, UTM

    Splines Interpolation– assignment in class

    Given the data:

     x 2 2.5 3 4

     f ( x) 5 7 8 2

    Predict f (3.4) using linear and quadratic splines.