S1 - Process product optimization using design experiments and response surface methodolgy

44
Process/product optimization using design of experiments and response surface methodology M. Mäkelä Sveriges landbruksuniversitet Swedish University of Agricultural Sciences Department of Forest Biomaterials and Technology Division of Biomass Technology and Chemistry Umeå, Sweden

description

An intensive practical course mainly for PhD-students on the use of designs of experiments (DOE) and response surface methodology (RSM) for optimization problems. The course covers relevant background, nomenclature and general theory of DOE and RSM modelling for factorial and optimisation designs in addition to practical exercises in Matlab. Due to time limitations, the course concentrates on linear and quadratic models on the k≤3 design dimension. This course is an ideal starting point for every experimental engineering wanting to work effectively, extract maximal information and predict the future behaviour of their system. Mikko Mäkelä (DSc, Tech) is a postdoctoral fellow at the Swedish University of Agricultural Sciences in Umeå, Sweden and is currently visiting the Department of Chemical Engineering at the University of Alicante. He is working in close cooperation with Paul Geladi, Professor of Chemometrics, and using DOE and RSM for process optimization mainly for the valorization of industrial wastes in laboratory and pilot scales.”

Transcript of S1 - Process product optimization using design experiments and response surface methodolgy

Page 1: S1 - Process product optimization using design experiments and response surface methodolgy

Process/product optimization using design of experiments and response surface methodology

M. Mäkelä

Sveriges landbruksuniversitetSwedish University of Agricultural Sciences

Department of Forest Biomaterials and TechnologyDivision of Biomass Technology and ChemistryUmeå, Sweden

Page 2: S1 - Process product optimization using design experiments and response surface methodolgy

DOE and RSM

DOE RSM

You

Design of experiments (DOE)

Planning experiments

→ Maximum information from

minimized number of experiments

Response Surface Methodology (RSM)

Identifying and fitting an appropriate

response surface model

→ Statistics, regression modelling &

optimization

Page 3: S1 - Process product optimization using design experiments and response surface methodolgy

What to expect?

Background and philosophy

Theory

Nomenclature

Practical demonstrations and exercises (Matlab)

What not?

Matrix algebra

Detailed equation studies

Statistical basics

Detailed listing of possible designs

Page 4: S1 - Process product optimization using design experiments and response surface methodolgy

Contents

Practical course, arranged in 4 individual sessions:

Session 1 – Introduction, factorial design, first order models

Session 2 – Matlab exercise: factorial design

Session 3 – Central composite designs, second order models, ANOVA,

blocking, qualitative factors

Session 4 – Matlab exercise: practical optimization example on given data

Page 5: S1 - Process product optimization using design experiments and response surface methodolgy

Session 1Introduction

Why experimental design

Factorial design

Design matrix

Model equation = coefficients

Residual

Response contour

Page 6: S1 - Process product optimization using design experiments and response surface methodolgy

If the current location is

known, a response surface

provides information on:

- Where to go

- How to get there

- Local maxima/minima

Response surfaces

Page 7: S1 - Process product optimization using design experiments and response surface methodolgy

Is there a difference?

vs. ?

Mäkelä et al., Appl. Energ. 131 (2014) 490.

Page 8: S1 - Process product optimization using design experiments and response surface methodolgy

Research problem

,

A and B constant reagents

C reaction product (response), to be maximized

T and P reaction conditions (continuous factors), can be regulated

Page 9: S1 - Process product optimization using design experiments and response surface methodolgy

Response as a contour plot

What kind of equation could

describe C behaviour as a

function of T and P?

C = f(T,P)

Page 10: S1 - Process product optimization using design experiments and response surface methodolgy

What else do we want to know?

Which factors and interactions are important

Positions of local optima (if they exist)

Surface and surface function around an

optimum

Direction towards an optimum

Statistical significance

Page 11: S1 - Process product optimization using design experiments and response surface methodolgy

How can we do it?

The expert method

Page 12: S1 - Process product optimization using design experiments and response surface methodolgy

How can we do it?

The shotgun method

Page 13: S1 - Process product optimization using design experiments and response surface methodolgy

How can we do it?

The ”Soviet” method

xk possibilities with k

factors on x levels

2 factors on 4 levels = 16

experiments

Page 14: S1 - Process product optimization using design experiments and response surface methodolgy

How can we do it?

The classical method

P fixed

T fixed

x

Page 15: S1 - Process product optimization using design experiments and response surface methodolgy

How can we do it?

Factorial design

∆T, ∆P

Factor interaction (diagonal)

Page 16: S1 - Process product optimization using design experiments and response surface methodolgy

Why experimental design?

Reduce the number of experiments

→ Cost, time

Extract maximal information

Understand what happens

Predict future behaviour

Page 17: S1 - Process product optimization using design experiments and response surface methodolgy

Challenges

Multiple factors on multiple levels

6 factors on 3 levels, 36 experiments

Reduce number of factors

Only 2 levels

→ Discard factors

= SCREENING

1

23

1

2

Page 18: S1 - Process product optimization using design experiments and response surface methodolgy

Factorial design

T

P

N:o T P

1 80 2

2 120 2

3 80 3

4 120 32

3

12080

Page 19: S1 - Process product optimization using design experiments and response surface methodolgy

Factorial design

T

P

1-1-1

1In coded levels

The smallest possible full factorial design!

N:o T T coded

P P coded

1 80 -1 2 -1

2 120 1 2 -1

3 80 -1 3 1

4 120 1 3 1

Page 20: S1 - Process product optimization using design experiments and response surface methodolgy

Factorial design

T

P

25 35

45 75

1-1-1

1Design matrix:

N:o T P C

1 -1 -1 25

2 1 -1 35

3 -1 1 45

4 1 1 75

Page 21: S1 - Process product optimization using design experiments and response surface methodolgy

Factorial design

T

P

25 35

45 75

1-1-1

1Average T effect:

T = 20

Average P effect:

P = 30

Interaction (TxP) effect:

TxP = 10

Page 22: S1 - Process product optimization using design experiments and response surface methodolgy

Research problem

, ,

A and B constant reagents

C reaction product (response), to be maximized

T, P and K reaction conditions (continuous factors) at two different levels

Number of experiments 23 = 9 ([levels][factors])

How to select proper factor levels?

Page 23: S1 - Process product optimization using design experiments and response surface methodolgy

Research problem

Empirical model:

, ,

In matrix notation:

yy⋮y

1 ⋯1 ⋯1 ⋮ ⋮ ⋱ ⋮1 ⋯

bb⋮b

ee⋮e

Measure ChooseUnknown!

Page 24: S1 - Process product optimization using design experiments and response surface methodolgy

Factorial design

First step

Selection and coding of factor levels

→ Design matrix

T = [80, 120]

P = [2, 3]

K = [0.5, 1]

0.5

280 120

3

1

P

T

K

Page 25: S1 - Process product optimization using design experiments and response surface methodolgy

Factorial design

Factorial design matrix

Notice symmetry in diffent colums

Inner product of two colums is zero

E.g. T’P = 0

This property is called orthogonality

N:o Order T P K

1 -1 -1 -1

2 1 -1 -1

3 -1 1 -1

4 1 1 -1

5 -1 -1 1

6 1 -1 1

7 -1 1 1

8 1 1 1

Randomize!

Page 26: S1 - Process product optimization using design experiments and response surface methodolgy

Orthogonality

For a first-order orthogonal design, X’X is a diagonal matrix:

If two columns are orthogonal, corresponding variables are linearly independent, i.e., assessed independent of each other.

1 11 11 11 1

, 1 1 1 11 1 1 1

1 1 1 11 1 1 1

1 11 11 11 1

4 00 4

2x4

4x22x2

Page 27: S1 - Process product optimization using design experiments and response surface methodolgy

Factorial design

N:o T P K Resp. (C)

1 -1 -1 -1 60

2 1 -1 -1 72

3 -1 1 -1 54

4 1 1 -1 68

5 -1 -1 1 52

6 1 -1 1 83

7 -1 1 1 45

8 1 1 1 80

Design matrix:

-1

-1-1 1

1

1

60 72

52 83

6854

45 80

T

PK

Page 28: S1 - Process product optimization using design experiments and response surface methodolgy

Factorial design

Model equation, main terms:

where

denotes response

factor (T, P or K)

coefficient

residual

mean term (average level)

N:o T P K Resp. (C)

1 -1 -1 -1 60

2 1 -1 -1 72

3 -1 1 -1 54

4 1 1 -1 68

5 -1 -1 1 52

6 1 -1 1 83

7 -1 1 1 45

8 1 1 1 80

Page 29: S1 - Process product optimization using design experiments and response surface methodolgy

Factorial design

Equation = coefficients

bbbb

64.211.52.50.8

bo average value (mean term) Large coefficient → important factor

Interactions usually present

Due to coding, the coefficients are comparable!

Page 30: S1 - Process product optimization using design experiments and response surface methodolgy

Factorial designModel equation with interactions:

N:o T P K TxP TxK PxK TxPxK Resp. (C)

1 -1 -1 -1 1 60

2 1 -1 -1 -1 72

3 -1 1 -1 1 54

4 1 1 -1 -1 68

5 -1 -1 1 -1 52

6 1 -1 1 1 83

7 -1 1 1 -1 45

8 1 1 1 1 80

Page 31: S1 - Process product optimization using design experiments and response surface methodolgy

Factorial design

+-

T

+

-P

+-

K

TxP

- +

TxK

+-

PxK

+

-

Main effects and interactions:

Page 32: S1 - Process product optimization using design experiments and response surface methodolgy

Factorial designEquation = coefficients

bbbbbbbb

64.211.52.50.80.85.000.3

Large interaction b13 (TxK)

Important interaction, main effects cannot be removed

→ Which coefficients to include?

Page 33: S1 - Process product optimization using design experiments and response surface methodolgy

Factorial design

An estimate of model error needed

Center-points

Duplicated experiments

Model residual

Page 34: S1 - Process product optimization using design experiments and response surface methodolgy

Factorial design

Error estimation allows significant testing

Remove insignificant coefficients

Leave main effects

Important interaction, main effect

cannot be removed

Page 35: S1 - Process product optimization using design experiments and response surface methodolgy

Factorial design

Error estimation allows significant testing

Remove insignificant coefficients

Leave main effects

Important interaction, main effect

cannot be removed

Recalculate significance upon removal!

Page 36: S1 - Process product optimization using design experiments and response surface methodolgy

Factorial design

Model residuals

Checking model adequacy

Finding outliers

Normally distributed

→ Random error

Several ways to present residuals

Possibility for response transformation

Page 37: S1 - Process product optimization using design experiments and response surface methodolgy

Factorial design

R2 statistic

Explained variability of

measured response

R2 = 0.9962

99.6% explained

Page 38: S1 - Process product optimization using design experiments and response surface methodolgy

Factorial design

More things to look at

Normal distribution of coefficients

Residual Standardized residual

Residual histogram

Residual vs. time

ANOVA

Page 39: S1 - Process product optimization using design experiments and response surface methodolgy

Factorial design

Page 40: S1 - Process product optimization using design experiments and response surface methodolgy

Factorial design

Prediction:

T = 110

K = 0.9

P = 2 (min. level)

Coded location:

1 0.5 1 0.6 0.3

Predicted response:

74.5 2.4

Page 41: S1 - Process product optimization using design experiments and response surface methodolgy

Session 1Introduction

Why experimental design

Factorial design

Design matrix

Model equation = coefficients

Residual

Response contour

Page 42: S1 - Process product optimization using design experiments and response surface methodolgy

Nomenclature

Factorial design

Screening

Design matrix

Model equation

Response

Effect (main/interaction)

Coefficient

Significance

Contour

Residual

Page 43: S1 - Process product optimization using design experiments and response surface methodolgy

Contents

Practical course, arranged in 4 individual sessions:

Session 1 – Introduction, factorial design, first order models

Session 2 – Matlab exercise: factorial design

Session 3 – Central composite designs, second order models, ANOVA,

blocking, qualitative factors

Session 4 – Matlab exercise: practical optimization example on given data

Page 44: S1 - Process product optimization using design experiments and response surface methodolgy

Thank you for listening!

Please send me an email that you are attending the course

[email protected]