EG 10111/10112 Introduction to Engineering Copyright © 2009 University of Notre Dame Probability and Limit States Introduction to Engineering Systems Lecture.

Slides:



Advertisements
Similar presentations
Design of Experiments Lecture I
Advertisements

EG 10111/10112 Introduction to Engineering Copyright © 2009 University of Notre Dame Introduction to Engineering Systems Lecture 3 (9/4/2009) Empirical.
6-1 Stats Unit 6 Sampling Distributions and Statistical Inference - 1 FPP Chapters 16-18, 20-21, 23 The Law of Averages (Ch 16) Box Models (Ch 16) Sampling.
Business Statistics for Managerial Decision
Analyzing Measurement Data ENGR 1181 Class 8. Analyzing Measurement Data in the Real World As previously mentioned, data is collected all of the time,
EG 10111/10112 Introduction to Engineering Copyright © 2009 University of Notre Dame Line fitting, Hooke’s Law, and Physics of Towers Introduction to Engineering.
EG 10111/10112 Introduction to Engineering Copyright © 2009 University of Notre Dame Technical Communication: Written documents Introduction to Engineering.
EG 10111/10112 Introduction to Engineering Copyright © 2009 University of Notre Dame Exam review (continued) Introduction to Engineering Systems Lecture.
EG 10111/10112 Introduction to Engineering Copyright © 2009 University of Notre Dame Design optimization: metrics and tower bracing Introduction to Engineering.
EG 10111/10112 Introduction to Engineering Copyright © 2009 University of Notre Dame Design optimization: optimization problem and factor of safety (F.O.S.)
EG 10111/10112 Introduction to Engineering Copyright © 2009 University of Notre Dame Statistics and Probability: Betting on a Design Introduction to Engineering.
EG 10111/10112 Introduction to Engineering Copyright © 2009 University of Notre Dame Exam review Introduction to Engineering Systems Lecture 11 (10/02/2009)
CHAPTER 6 Statistical Analysis of Experimental Data
Section The Idea of Probability Statistics.
QMS 6351 Statistics and Research Methods Probability and Probability distributions Chapter 4, page 161 Chapter 5 (5.1) Chapter 6 (6.2) Prof. Vera Adamchik.
Lecture 6: Descriptive Statistics: Probability, Distribution, Univariate Data.
Chapter 16: Random Variables
CHAPTER 10: Introducing Probability
Engineering Probability and Statistics
 A number expressing the likelihood that a specific event will occur, expressed as the ratio of the number of actual occurrences.
Math 647 March 18 & 20, Probability Probability: Probability: Probability of exceedance: Pr[ X > x ] or Pr [X x ] or Pr [X < x ] Relative frequency/proportions:
Sullivan – Fundamentals of Statistics – 2 nd Edition – Chapter 5 Section 1 – Slide 1 of 33 Chapter 5 Section 1 Probability Rules.
Theory of Probability Statistics for Business and Economics.
Applied Business Forecasting and Regression Analysis Review lecture 2 Randomness and Probability.
● Uncertainties abound in life. (e.g. What's the gas price going to be next week? Is your lottery ticket going to win the jackpot? What's the presidential.
Probability Distributions. Essential Question: What is a probability distribution and how is it displayed?
Copyright © 2006 Brooks/Cole, a division of Thomson Learning, Inc.
OPIM 5103-Lecture #3 Jose M. Cruz Assistant Professor.
LECTURER PROF.Dr. DEMIR BAYKA AUTOMOTIVE ENGINEERING LABORATORY I.
Outline Random processes Random variables Probability histograms
1 Chapter 4. Section 4-1 and 4-2. Triola, Elementary Statistics, Eighth Edition. Copyright Addison Wesley Longman M ARIO F. T RIOLA E IGHTH E DITION.
P. STATISTICS LESSON 8.2 ( DAY 1 )
Binomial Experiment A binomial experiment (also known as a Bernoulli trial) is a statistical experiment that has the following properties:
Psyc 235: Introduction to Statistics DON’T FORGET TO SIGN IN FOR CREDIT!
1.3 Simulations and Experimental Probability (Textbook Section 4.1)
Chapter 16: Random Variables
LECTURE 3: ANALYSIS OF EXPERIMENTAL DATA
CHAPTER 10: Introducing Probability ESSENTIAL STATISTICS Second Edition David S. Moore, William I. Notz, and Michael A. Fligner Lecture Presentation.
NLP. Introduction to NLP Very important for language processing Example in speech recognition: –“recognize speech” vs “wreck a nice beach” Example in.
40S Applied Math Mr. Knight – Killarney School Slide 1 Unit: Statistics Lesson: ST-5 The Binomial Distribution The Binomial Distribution Learning Outcome.
Chapter 10 Introducing Probability BPS - 5th Ed. Chapter 101.
Sections 5.1 and 5.2 Review and Preview and Random Variables.
L56 – Discrete Random Variables, Distributions & Expected Values
1 Keep Life Simple! We live and work and dream, Each has his little scheme, Sometimes we laugh; sometimes we cry, And thus the days go by.
MATH 256 Probability and Random Processes Yrd. Doç. Dr. Didem Kivanc Tureli 14/10/2011Lecture 3 OKAN UNIVERSITY.
CHAPTER 10: Introducing Probability ESSENTIAL STATISTICS Second Edition David S. Moore, William I. Notz, and Michael A. Fligner Lecture Presentation.
BPS - 5th Ed. Chapter 101 Introducing Probability.
Unit 4: Probability Day 2: Basic Probability. Standards and Benchmarks Select and apply counting procedures, such as the multiplication and addition.
S ECTION 5.3 – S IMULATIONS. W HAT IS A SIMULATION ? A simulation is a mock trial of an experiment without doing the experiment. It uses theoretical probabilities.
Chapter 2: Probability. Section 2.1: Basic Ideas Definition: An experiment is a process that results in an outcome that cannot be predicted in advance.
Chapter 8: Probability: The Mathematics of Chance Probability Models and Rules 1 Probability Theory  The mathematical description of randomness.  Companies.
How do we classify uncertainties? What are their sources? – Lack of knowledge vs. variability. What type of measures do we take to reduce uncertainty?
Section The Idea of Probability AP Statistics
Week 5 Discrete Random Variables and Probability Distributions Statistics for Social Sciences.
Probability & Random Variables
1 ENGINEERING MEASUREMENTS Prof. Emin Korkut. 2 Statistical Methods in Measurements.
Statistics 11/7/ Statistics: Normal Curve CSCE 115.
Theoretical distributions: the Normal distribution.
Random Error Theory.
MECH 373 Instrumentation and Measurements
Statistical Methods Michael J. Watts
Sec. 4-5: Applying Ratios to Probability
Chapter 6 6.1/6.2 Probability Probability is the branch of mathematics that describes the pattern of chance outcomes.
Statistical Methods Michael J. Watts
CHAPTER 6 Random Variables
The Law of Large Numbers
What are models and simulations?
CHAPTER 10: Introducing Probability
Section 6.1 The Idea of Probability
M248: Analyzing data Block A UNIT A3 Modeling Variation.
Presentation transcript:

EG 10111/10112 Introduction to Engineering Copyright © 2009 University of Notre Dame Probability and Limit States Introduction to Engineering Systems Lecture 6 (9/14/2009) Prof. Andrés Tovar

EG 10111/10112 Introduction to Engineering Copyright © 2009 University of Notre Dame Announcements Print out Learning Center documents, read them, and bring them to Learning Center There’s a typo in HW2 Betting on a Design2 should be (1.25, 27)

EG 10111/10112 Introduction to Engineering Copyright © 2009 University of Notre Dame Note on polyfit Note that polyfit only finds coefficients of line (slope and intercept). It does not plot the line. –need to add plot command to do this Betting on a Design3 x = 0:.2:1; y_exp = [ ]; c = polyfit(x, y_exp, 1); m = c(1); % slope b = c(2); % intercept y_fit = m*x + b; plot(x, y_exp, 'o') % plot experimental data hold on plot(x, y_fit, 'r') % plot best-fit line hold off

EG 10111/10112 Introduction to Engineering Copyright © 2009 University of Notre Dame From last class Basis function determines the empirical model. Statistic is a numerical value that characterizes a dataset. –Mean or average (see also median and mode) –Standard deviation –Coefficient of variation Uncertainty can be epistemic or aleatory. Error can be systematic or random. Betting on a Design4

EG 10111/10112 Introduction to Engineering Copyright © 2009 University of Notre Dame Firing at a Target Adjust the pullback distance X of the slingshot to hit a target at a distance D downrange. How likely is it that a softball launched with a pullback of 1 m will land within 1 m of 17 m? Betting on a Design5  = 17.2 m  = 1.4 m

EG 10111/10112 Introduction to Engineering Copyright © 2009 University of Notre Dame Where to Aim the Slingshot? Suppose that you are required to take the following bet: –You will be paid $1 to launch a ball to less than 18 m. –But you will be fined $5 for each launch shorter than 16 m What target distance should you aim for to maximize your gain? –How would you solve it? What would you do differently if the fine were increased to $50? Betting on a Design win $1 lose $5 6

EG 10111/10112 Introduction to Engineering Copyright © 2009 University of Notre Dame Probability The relative frequency of possible outcomes It is measured on a scale from 0 to 1 Estimate probability by running an experiment with multiple trials –and counting number of favorable/unfavorable outcomes Betting on a Design estimated experimental probability of winning number of successful outcomes total number of trials = probability of losing1 – probability of winning = 7

EG 10111/10112 Introduction to Engineering Copyright © 2009 University of Notre Dame Coin toss Suppose we run an experiment where we toss a coin 3 times H, H, T What are the possible probabilities of the coin coming up “heads” that we could estimate from this experiment? 2/3 ≈ = 66.67% What effect would using more trials have on our estimate? 1/2 = 0.5 = 50% Betting on a Design8

EG 10111/10112 Introduction to Engineering Copyright © 2009 University of Notre Dame More games Betting on a Design9 1/36 Theoretical probability of single events

EG 10111/10112 Introduction to Engineering Copyright © 2009 University of Notre Dame Example: Rolling a Die Given a fair 6-sided die, what is your expected gain if: –you win $5 if you roll 1 or 6 –you lose $2 if you roll 2, 3, 4, or 5 Should you take this bet? Betting on a Design10 % simulation n = ; r = randi(6,1,n); wins = sum(r==1|r==6); losses = sum(r==2|r==3|r==4|r==5); simulated_gain = (5*wins - 2*losses)/n % theory p = 2; a = 5; q = 4; b = -2; theoretical_gain = (p*a + q*b)/(p + q) %#ok

EG 10111/10112 Introduction to Engineering Copyright © 2009 University of Notre Dame Expected Gain (or Loss) From a Bet Huygens (again), 1656 Given – p chances of winning a sum of money a – q chances of winning a sum of money b Betting on a Design expected gain 11

EG 10111/10112 Introduction to Engineering Copyright © 2009 University of Notre Dame Estimating Probabilities for the Slingshot For firing a slingshot with a pullback of 1 m. Estimate the probability that: –ball lands within 1 m of the target at 17 m. –ball lands short of this range (>18 m) –ball lands beyond this range (<16 m) Betting on a Design12

EG 10111/10112 Introduction to Engineering Copyright © 2009 University of Notre Dame Bin Counts Betting on a Design13

EG 10111/10112 Introduction to Engineering Copyright © 2009 University of Notre Dame Histogram Betting on a Design14

EG 10111/10112 Introduction to Engineering Copyright © 2009 University of Notre Dame Probabilities Betting on a Design15

EG 10111/10112 Introduction to Engineering Copyright © 2009 University of Notre Dame Theory of the Bell Curve: Normal Distribution Betting on a Design16 flips from concave up to concave down 2.1% 13.6% 2.1%0.1% 99.7% of distribution is between  3  99.9% of distribution is less than  + 3 

EG 10111/10112 Introduction to Engineering Copyright © 2009 University of Notre Dame Experimental vs. theoretical distribution Betting on a Design17 % distances of the softball D = [ ]; m = mean(D); % mean s = std(D); % standard deviation % experimental probability density y = 14:1:20; n = histc(D,y); bar(y+1/2,n/length(D),1); hold on xlabel('distance d (m)') ylabel('probability p(d)') % normal probability density itv = ; x = m-3*s:itv:m+3*s; p = 1/(s*sqrt(2*pi))*exp(-(x-m).^2/(2*s^2)); plot(x,p,'r.'); hold off p_theo=sum(p(x>=16 & x<=18))*itv p(16≤d≤18) = 0.55 (experimental) p(16≤d≤18) = 0.53 (normal distribution)

EG 10111/10112 Introduction to Engineering Copyright © 2009 University of Notre Dame The probabilities Betting on a Design18 % distances of the softball D = [ ]; m = mean(D); % mean s = std(D); % standard deviation % plots itv = ; x = (m-3*s:itv:m+3*s); px = pdf('norm',m-3*s:itv:m+3*s,m,s); % probability density function figure(1); plot(x,px,'r.') % cumulative distribution function pxc = cdf('norm',m-3*s:itv:m+3*s,m,s); figure(2); plot(x,pxc,'b.') % results px16 = cdf('norm',16,m,s) % x<=16 pxm = cdf('norm',16:18,m,s); px17 = pxm(3)-pxm(1) % 16<x<=18 px18 = 1-cdf('norm',18,m,s) % x>18

EG 10111/10112 Introduction to Engineering Copyright © 2009 University of Notre Dame Should we take the bet? Betting on a Design19 p = , a = 1 q = , b = -5

EG 10111/10112 Introduction to Engineering Copyright © 2009 University of Notre Dame Where we are in the tower design process Betting on a Design20 Gather Data Develop Model Verify Model MODEL DEVELOPMENT Investigate Designs using Model Optimize Design Predict Behavior DESIGN STAGE Construct Design Experimentally Verify Behavior CONSTRUCTION & VERIFICATION

EG 10111/10112 Introduction to Engineering Copyright © 2009 University of Notre Dame Tower Bracing Betting on a Design21

EG 10111/10112 Introduction to Engineering Copyright © 2009 University of Notre Dame Design Considerations Betting on a Design22

EG 10111/10112 Introduction to Engineering Copyright © 2009 University of Notre Dame Limit States Limit state: objective on the performance or behavior of a design The consequences of failing to meet limit states can be very high –customer dissatisfaction (loss of business) –product recall –loss of property –loss of life Terminology (“limit state design”) comes from structural engineering, but the concept cuts across all fields Betting on a Design23

EG 10111/10112 Introduction to Engineering Copyright © 2009 University of Notre Dame Limit States for Skyscraper Design Betting on a Design24

EG 10111/10112 Introduction to Engineering Copyright © 2009 University of Notre Dame Distribution of Tower Stiffness Betting on a Design N/mm

EG 10111/10112 Introduction to Engineering Copyright © 2009 University of Notre Dame Histogram of Tower Trial Stiffnesses Betting on a Design26  = 0.07 N/mm  = 0.02 N/mm

EG 10111/10112 Introduction to Engineering Copyright © 2009 University of Notre Dame How Good of a Fit for Tower Stiffnesses? Pretty good fit! –We can use the normal distribution model to estimate probabilities Betting on a Design27  = 0.07  = 0.02

EG 10111/10112 Introduction to Engineering Copyright © 2009 University of Notre Dame Where to Aim for Tower Stiffness Betting on a Design28 fail pass limit state = 4.5N/19mm bracing cost expensive cheap

EG 10111/10112 Introduction to Engineering Copyright © 2009 University of Notre Dame Confidence in Meeting Limit State Use probabilities from normal distribution to determine where to aim your design 29 Confidence in exceeding limit state Probability of Failure Aim design at 50.0% limit state value 84.2%15.8% limit state value +  97.8%2.2% limit state value + 2  99.9%0.1% limit state value + 3  Betting on a Design

EG 10111/10112 Introduction to Engineering Copyright © 2009 University of Notre Dame This week The tower design problem –Factor of safety –Constraints –Efficiency E = S/B S = k br /k ubr : stiffness ratio B = L br /L ubr : bracing ratio Introduction to design optimization Betting on a Design30