49
Propagation of Uncertainty Jake Blanchard Spring 2010 Uncertainty Analysis for Engineers 1

Propagation of Uncertainty

  • Upload
    lazar

  • View
    48

  • Download
    1

Embed Size (px)

DESCRIPTION

Propagation of Uncertainty. Jake Blanchard Spring 2010. Introduction. We’ve discussed single-variable probability distributions This lets us represent uncertain inputs But what of variables that depend on these inputs? How do we represent their uncertainty? - PowerPoint PPT Presentation

Citation preview

Page 1: Propagation of Uncertainty

Uncertainty Analysis for Engineers 1

Propagation of UncertaintyJake BlanchardSpring 2010

Page 2: Propagation of Uncertainty

Uncertainty Analysis for Engineers 2

IntroductionWe’ve discussed single-variable

probability distributionsThis lets us represent uncertain

inputsBut what of variables that

depend on these inputs? How do we represent their uncertainty?

Some problems can be done analytically; others can only be done numerically

These slides discuss analytical approaches

Page 3: Propagation of Uncertainty

Uncertainty Analysis for Engineers 3

Functions of 1 Random VariableSuppose we have Y=g(X) where

X is a random input variableAssume the pdf of X is

represented by fx.If this pdf is discrete, then we can

just map pdf of X onto YIn other words X=g-1(Y)So fy(Y)=fx[g-1(y)]

Page 4: Propagation of Uncertainty

Uncertainty Analysis for Engineers 4

ExampleConsider Y=X2.Also, assume discrete pdf of X is

as shown belowWhen X=1, Y=1; X=2, Y=4; X=3,

Y=9

0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.50

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0 5 10 15 20 25 300

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

Page 5: Propagation of Uncertainty

Uncertainty Analysis for Engineers 5

Discrete VariablesExample:

◦Manufacturer incurs warranty charges for system breakdowns

◦Charge is C for the first breakdown, C2 for the second failure, and Cx for the xth breakdown (C>1)

◦Time between failures is exponentially distributed (parameter ), so number of failures in period T is Poisson variate with parameter T

◦What is distribution for warranty cost for T=1 year

Page 6: Propagation of Uncertainty

Uncertainty Analysis for Engineers 6

Formulation

...,,!)ln(

)ln(

0)(

...,,)ln()ln(

00

...,2,100

)(

...,2,1,0!

)(

2)ln(

)ln(

2

CCw

Cw

e

wewp

CCwCw

wx

xCx

xhw

xxexf

Cw

x

x

Page 7: Propagation of Uncertainty

Uncertainty Analysis for Engineers 7

Plots

C=2=1

1 1.5 2 2.5 3 3.5 4 4.5 50

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

x

0 5 10 15 20 25 30 350

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

w

Page 8: Propagation of Uncertainty

Uncertainty Analysis for Engineers 8

CDF For Discrete DistributionsIf g(x) monotonically increases,

then P(Y<y)=P[X<g-1(y)]If g(x) monotonically decreases,

then P(Y<y)=P[X>g-1(y)]…and, formally,

)(

1

1

)()()(ygx

ixXYi

xpygFyF

x

y

x

y

Page 9: Propagation of Uncertainty

Uncertainty Analysis for Engineers 9

Another ExampleSuppose Y=X2 and X is Poisson

with parameter

,...9,4,1,0

!

,...3,2,1,0!

)(

)(1

2

yeytp

xextp

YXYg

XgXY

ty

y

tx

x

Page 10: Propagation of Uncertainty

Uncertainty Analysis for Engineers 10

Continuous DistributionsIf fx is continuous, it takes a bit

more work

dydggf

dydFyf

or

dydy

ydgygfyF

dydy

ydgdx

ygx

dxxfdxxfyF

xy

y

xY

yg

xygxxY

11

11

1

1

)(

)(

)(

)()(

)(

)()()(1

1

Page 11: Propagation of Uncertainty

Uncertainty Analysis for Engineers 11

Example

2exp

21

2exp

21

21exp

21

)(

2

2

2

1

1

yf

yf

Xf

imaginedydg

YygX

XY

y

y

x

Normal distribution

Mean=0, =1

Page 12: Propagation of Uncertainty

Uncertainty Analysis for Engineers 12

ExampleX is

lognormal

2

2

2

1

1

21exp

21

)exp(21exp

)exp(21

)ln(21exp

21

)exp(

)exp()(

)ln(

yf

yyy

f

xx

f

imagine

Ydydg

YygX

XY

y

y

x

Normal distributi

on

Page 13: Propagation of Uncertainty

Uncertainty Analysis for Engineers 13

If g-1(y) is multi-valued…

),(

21

21

)(

2

1

11

ognormallS

cucuf

cuff

cududS

cuS

cSU

Exampledydggfyf

ssu

k

iixY

Page 14: Propagation of Uncertainty

Uncertainty Analysis for Engineers 14

Example (continued)

22ln

22lnln

21exp

221

21

ln

21exp

2

1

)ln(21exp

21

2

2

2

u

u

u

u

s

c

cuu

f

cu

cu

cu

f

ss

f

lognormal

Page 15: Propagation of Uncertainty

Uncertainty Analysis for Engineers 15

Example

00

00

22

exp121

21

21

21

0exp1

1400

vaz

vaz

azazf

azazf

azff

azdzdV

aZv

vvv

vf

imagine

aVd

FVZ

vvvz

v

Page 16: Propagation of Uncertainty

Uncertainty Analysis for Engineers 16

A second exampleSuppose we are making strips of

sheet metalIf there is a flaw in the sheet, we must

discard some materialWe want an assessment of how much

waste we expectAssume flaws lie in line segments (of

constant length L) making an angle with the sides of the sheet

is uniformly distributed from 0 to

Page 17: Propagation of Uncertainty

Uncertainty Analysis for Engineers 17

Schematic

L

w

Page 18: Propagation of Uncertainty

Uncertainty Analysis for Engineers 18

Example (continued)Whenever a flaw is found, we

must cut out a segment of width w

22

2/12

1

111

sin

,0sin

wLLw

Ldwd

Lww

UfLhw

Page 19: Propagation of Uncertainty

Uncertainty Analysis for Engineers 19

Example (continued)g-1 is multi-

valued

2221

222

221

2

01

01

wLwfwff

LwwL

wf

LwwL

wf

w

</2

>/2

Page 20: Propagation of Uncertainty

Uncertainty Analysis for Engineers 20

Results

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 10.5

1

1.5

2

2.5

3

3.5

4

4.5

5

wL=1

cdf

pdf

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 10

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

w

Page 21: Propagation of Uncertainty

Uncertainty Analysis for Engineers 21

Functions of Multiple Random VariablesZ=g(X,Y)For discrete variables

If we have the sum of random variables

Z=X+Y iji xall

iiyxzyx

jiyxz xzxfyxff ,),( ,,

zyxg

jiyxzji

yxff),(

, ),(

Page 22: Propagation of Uncertainty

Uncertainty Analysis for Engineers 22

ExampleZ=X+Y

0.5 1 1.5 2 2.5 3 3.50

0.10.20.30.40.50.60.7

x

fx

5 10 15 20 25 30 350

0.050.1

0.150.2

0.250.3

0.350.4

0.45

y

fy

Page 23: Propagation of Uncertainty

Uncertainty Analysis for Engineers 23

AnalysisX Y Z P Z-rank1 10 11 .08 11 20 21 .04 41 30 31 .08 72 10 12 .24 22 20 22 .12 52 30 32 .24 83 10 13 .08 33 20 23 .04 63 30 33 .08 9

Page 24: Propagation of Uncertainty

Uncertainty Analysis for Engineers 24

Result

5 10 15 20 25 30 350

0.05

0.1

0.15

0.2

0.25

0.3

Z

fz

Page 25: Propagation of Uncertainty

Uncertainty Analysis for Engineers 25

ExampleZ=X+Y

0.5 1 1.5 2 2.5 3 3.50

0.10.20.30.40.50.60.7

x

fx

1.5 2 2.5 3 3.5 4 4.50

0.050.1

0.150.2

0.250.3

0.350.4

0.45

fy

y

Page 26: Propagation of Uncertainty

Uncertainty Analysis for Engineers 26

AnalysisX Y Z P Z-rank1 2 3 .08 11 3 4 .04 21 4 5 .08 32 2 4 .24 22 3 5 .12 32 4 6 .24 43 2 5 .08 33 3 6 .04 43 4 7 .08 5

Page 27: Propagation of Uncertainty

Uncertainty Analysis for Engineers 27

Compiled Dataz fz3 .084 .285 .286 .287 .08

2.5 3 3.5 4 4.5 5 5.5 6 6.5 7 7.50

0.050.1

0.150.2

0.250.3

fz

z

Page 28: Propagation of Uncertainty

Uncertainty Analysis for Engineers 28

Example

allx

xzxz

z

allx

xzx

xallyxz

y

y

x

x

xzxvtvtf

tvxzxtvtxzfxff

tytf

vtxvtf

YXZ

)!(!exp

exp)!(!

)()(

)exp(!

)exp(!

x and y are integers

Page 29: Propagation of Uncertainty

Uncertainty Analysis for Engineers 29

Example (continued)

tvz

tvf

zxzxv

z

z

z

allx

xzx

exp!

!)!(!

The sum of n independent Poisson processes is Poisson

Page 30: Propagation of Uncertainty

Uncertainty Analysis for Engineers 30

Continuous Variables

z

yxz

z

yxz

g

yxz

zyxgyxz

dydzdzdgygfzF

dzdydzdgygfzF

gyzgx

dxdyyxfzF

dxdyyxfzF

YXgZ

11

,

11

,

11

,

),(,

),()(

),()(

),(

),()(

),()(

),(

1

Page 31: Propagation of Uncertainty

Uncertainty Analysis for Engineers 31

Continuous Variables

dYYabYZf

af

adzdg

abYZX

bYaXZif

dXdzdggXf

dYdzdgYgfzf

yxz

yx

yxz

),(1

1

),(

),()(

,

1

11

,

11

,

Page 32: Propagation of Uncertainty

Uncertainty Analysis for Engineers 32

Continuous Variables (cont.)

dyYfabYZf

af

tindependenyx

dXbaXZXf

bf

dYYabYZf

af

yxz

yxz

yxz

)()(1

,

),(1

),(1

,

,

Page 33: Propagation of Uncertainty

Uncertainty Analysis for Engineers 33

Example

mw

mf

uwudu

mw

mf

duuwum

wm

f

dumuw

uwmmu

muduuwfuff

mv

mvf

mu

muf

VUVUW

w

w

w

w

vuw

v

u

2exp

21

)(2exp

21

112

exp21

2)(exp

)(21

2exp

21)()(

2exp

21

2exp

21

0,;

0

Page 34: Propagation of Uncertainty

Uncertainty Analysis for Engineers 34

In General…If Z=X+Y and X and Y are normal dist.

Then Z is also normal with222yxz

yxz

2

2222

22

22

21exp

2

1

21

21exp

21

21exp

21

21exp

21

)()(

yx

yx

yx

z

y

y

x

x

yxz

y

y

yx

x

xz

yxz

zf

dyyyzf

dyyyzf

dyyfyzff

Page 35: Propagation of Uncertainty

Uncertainty Analysis for Engineers 35

Products

n

iXZ

n

iXZ

n

ii

n

ii

i

YXz

i

i

XZ

XZ

andognormallallXimagine

dyyyzf

YZf

YZX

YZX

XYZ

1

22

1

1

1

,

)ln()ln(

),(1)(

1

Page 36: Propagation of Uncertainty

Uncertainty Analysis for Engineers 36

ExampleW, F, E are lognormal

2222

2121

)ln(21)ln()ln()ln(

EFWC

EFWC

EFWC

EWFC

Page 37: Propagation of Uncertainty

Uncertainty Analysis for Engineers 37

Central Limit TheoremThe sum of a large number of

individual random components, none of which is dominant, tends to the Gaussian distribution (for large n)

Page 38: Propagation of Uncertainty

Uncertainty Analysis for Engineers 38

GeneralizationMore than two variables…

nnxxZ

n

dxdxdxzgxxxgfzf

xxxxgZ

n...,...,,,...)(

),...,,,(

32

1

321

,...,

321

1

Page 39: Propagation of Uncertainty

Uncertainty Analysis for Engineers 39

MomentsSuppose Z=g(X1, X2, …,Xn)

bXaEdxxfbdxxfxaYE

dxxfbaxdyyfYYE

baXYimagine

dXdXdXXXXfXXXgZE

dXdXdXXXXfzZE

xx

xy

nnXXXn

nnXXX

n

n

)()()()(

)()()()(

...),...,,(),...,,(...)(

...),...,,(...)(

2121,...,,21

2121,...,,

21

21

Page 40: Propagation of Uncertainty

Uncertainty Analysis for Engineers 40

Moments

)()(

)()()(

)()()(

)()(

2

22

2

22

XVaraYVar

dxxfxExaYVar

dxxfbxaEbaxYVar

dxxfbaxYEYVar

x

x

xYY

Page 41: Propagation of Uncertainty

Uncertainty Analysis for Engineers 41

Moments

)()()(),(),(2)()()(

),(2

),(

),()(

),()(

),()(

)()()(

21212121

2122

12

2121,21

2121,2

22

2121,2

12

2121,2

21

2121,2

21

21

21

2121

212

211

2121

21

XEXEXXEXXEXXCovXXabCovXVarbXVaraYVar

dxdxxxfxxab

dxdxxxfxb

dxdxxxfxaYVar

dxdxxxfbabxaxYVar

dxdxxxfyYVar

XbEXaEYEbXaXY

imagine

XX

xxxx

xxx

xxx

xxxx

xxy

Page 42: Propagation of Uncertainty

Uncertainty Analysis for Engineers 42

Approximation

)()(

)()()()(

)()()()(

)()()(

)()(

)()()(

)(

x

xxxx

xxxx

xxx

xx

x

gYE

dXXfXdxdgdXXfgYE

dXXfdxdgXdXXfgYE

dXXfdxdgXgYE

dxdgXgXg

dXXfXgYE

XgY

Page 43: Propagation of Uncertainty

Uncertainty Analysis for Engineers 43

Approximation

2

22

2

2

2

)(

)()(

)()(

)()()(

)()(

)()()(

xx

xx

xx

xyxx

xx

xy

dxdgXVarYVar

dXXfXdxdgYVar

dXXfdxdgXYVar

dXXfdxdgXgYVar

dxdgXgXg

dXXfXgYVar

Page 44: Propagation of Uncertainty

Uncertainty Analysis for Engineers 44

Second Order Approximation

)(21)()(

)(21)()(

)(21)()(

21)()(

)()()(

)(

2

2

22

2

2

22

2

22

xVardxgdgYE

dXXfXdxgdgYE

dXXfdxgdXgYE

dxgdX

dxdgXgXg

dXXfXgYE

XgY

xxx

xxx

xxx

xxx

x

Page 45: Propagation of Uncertainty

Uncertainty Analysis for Engineers 45

Approximation for Multiple Inputs

n

i iX

i

n

ixXXXX

n

XgYVar

xggYE

XXXXgY

i

in

1

22

2

2

1

2

321

)(

21,...,,,)(

),...,,,(

321

Page 46: Propagation of Uncertainty

Uncertainty Analysis for Engineers 46

ExampleExample 4.13Do exact and then use

approximation and compareWaste Treatment Plant – C=cost,

W=weight of waste, F=unit cost factor, E=efficiency coefficient

median covW 2000 ton/y .2F $20/ton .15E 1.6 .125

EWFC

Page 47: Propagation of Uncertainty

Uncertainty Analysis for Engineers 47

Solving…

84771exp

620,3221exp

25563.041

36.1021ln

ln21lnlnln

124516.0cov1ln

149166.0cov1ln

19804.0cov1ln

4700.0ln9957.2ln6009.7ln

2

2

222

2

2

2

CCC

CCC

EFWC

EFWC

EE

FF

WW

medianE

medianF

medianW

CE

EFWC

EFW

Page 48: Propagation of Uncertainty

Uncertainty Analysis for Engineers 48

Approximation

%16.032620

3262032673

3267321

43

0

%4.032620

3248332620;483,32

2016.0;033.3;915.4076124.1;223.20;6.2039

,,

21

21

21,,)(

2

22

2/52

2

2

2

2

2

2

22

2

22

2

22

error

Eg

Eg

Fg

Wg

error

g

Eg

Fg

WggCE

EE

FW

E

FW

E

FW

EFW

EFW

E

FWEFW

EFWEFW

Page 49: Propagation of Uncertainty

Uncertainty Analysis for Engineers 49

Variance

%3.1847783708477

8370

2)(

)(

2

2/32

2

2

2

2

22

22

22

error

CVar

Eg

Fg

WgCVar

C

E

FWE

E

WF

E

FW

EFW