University of Colorado Boulder ASEN 5070: Statistical Orbit Determination I Fall 2015 Professor Brandon A. Jones Lecture 19: Examples with the Batch Processor.

Slides:



Advertisements
Similar presentations
Modeling of Data. Basic Bayes theorem Bayes theorem relates the conditional probabilities of two events A, and B: A might be a hypothesis and B might.
Advertisements

The Normal Distribution
Kin 304 Regression Linear Regression Least Sum of Squares
Copyright © 2006 The McGraw-Hill Companies, Inc. Permission required for reproduction or display. 1 ~ Curve Fitting ~ Least Squares Regression Chapter.
University of Colorado Boulder ASEN 5070: Statistical Orbit Determination I Fall 2014 Professor Brandon A. Jones Lecture 28: Orthogonal Transformations.
Chapter 10 Curve Fitting and Regression Analysis
Use of regression analysis Regression analysis: –relation between dependent variable Y and one or more independent variables Xi Use of regression model.
Simple Regression. Major Questions Given an economic model involving a relationship between two economic variables, how do we go about specifying the.
University of Colorado Boulder ASEN 5070: Statistical Orbit Determination I Fall 2014 Professor Brandon A. Jones Lecture 24: Numeric Considerations and.
University of Colorado Boulder ASEN 5070: Statistical Orbit Determination I Fall 2014 Professor Brandon A. Jones Lecture 20: Project Discussion and the.
Observers and Kalman Filters
University of Colorado Boulder ASEN 5070 Statistical Orbit determination I Fall 2012 Professor George H. Born Professor Jeffrey S. Parker Lecture 8: Stat.
Lecture 9 Inexact Theories. Syllabus Lecture 01Describing Inverse Problems Lecture 02Probability and Measurement Error, Part 1 Lecture 03Probability and.
The Simple Linear Regression Model: Specification and Estimation
Environmentally Conscious Design & Manufacturing (ME592) Date: May 5, 2000 Slide:1 Environmentally Conscious Design & Manufacturing Class 25: Probability.
Probability & Statistics for Engineers & Scientists, by Walpole, Myers, Myers & Ye ~ Chapter 11 Notes Class notes for ISE 201 San Jose State University.
University of Colorado Boulder ASEN 5070: Statistical Orbit Determination I Fall 2014 Professor Brandon A. Jones Lecture 7: Linearization and the State.
Variance and covariance Sums of squares General linear models.
University of Colorado Boulder ASEN 5070: Statistical Orbit Determination I Fall 2014 Professor Brandon A. Jones Lecture 37: SNC Example and Solution Characterization.
Principles of the Global Positioning System Lecture 11 Prof. Thomas Herring Room A;
Colorado Center for Astrodynamics Research The University of Colorado STATISTICAL ORBIT DETERMINATION Project Report Unscented kalman Filter Information.
University of Colorado Boulder ASEN 5070 Statistical Orbit Determination I Fall 2012 Professor Jeffrey S. Parker Professor George H. Born Lecture 25: Error.
Chapter 15 Modeling of Data. Statistics of Data Mean (or average): Variance: Median: a value x j such that half of the data are bigger than it, and half.
Stats for Engineers Lecture 9. Summary From Last Time Confidence Intervals for the mean t-tables Q Student t-distribution.
University of Colorado Boulder ASEN 5070 Statistical Orbit determination I Fall 2012 Professor George H. Born Professor Jeffrey S. Parker Lecture 7: Spaceflight.
Modern Navigation Thomas Herring
University of Colorado Boulder ASEN 5070 Statistical Orbit determination I Fall 2012 Professor George H. Born Professor Jeffrey S. Parker Lecture 5: Stat.
University of Colorado Boulder ASEN 5070: Statistical Orbit Determination I Fall 2014 Professor Brandon A. Jones Lecture 18: Minimum Variance Estimator.
University of Colorado Boulder ASEN 5070: Statistical Orbit Determination I Fall 2014 Professor Brandon A. Jones Lecture 14: Probability Wrap-Up and Statistical.
University of Colorado Boulder ASEN 5070: Statistical Orbit Determination I Fall 2014 Professor Brandon A. Jones Lecture 26: Singular Value Decomposition.
Colorado Center for Astrodynamics Research The University of Colorado 1 STATISTICAL ORBIT DETERMINATION ASEN 5070 LECTURE 11 9/16,18/09.
University of Colorado Boulder ASEN 5070: Statistical Orbit Determination I Fall 2015 Professor Brandon A. Jones Lecture 21: A Bayesian Approach to the.
Colorado Center for Astrodynamics Research The University of Colorado 1 STATISTICAL ORBIT DETERMINATION The Minimum Variance Estimate ASEN 5070 LECTURE.
University of Colorado Boulder ASEN 5070 Statistical Orbit determination I Fall 2012 Professor George H. Born Professor Jeffrey S. Parker Lecture 11: Batch.
Geology 5670/6670 Inverse Theory 21 Jan 2015 © A.R. Lowry 2015 Read for Fri 23 Jan: Menke Ch 3 (39-68) Last time: Ordinary Least Squares Inversion Ordinary.
Colorado Center for Astrodynamics Research The University of Colorado 1 STATISTICAL ORBIT DETERMINATION Probability and statistics review ASEN 5070 LECTURE.
Economics 173 Business Statistics Lecture 10 Fall, 2001 Professor J. Petry
NON-LINEAR REGRESSION Introduction Section 0 Lecture 1 Slide 1 Lecture 6 Slide 1 INTRODUCTION TO Modern Physics PHYX 2710 Fall 2004 Intermediate 3870 Fall.
University of Colorado Boulder ASEN 5070: Statistical Orbit Determination I Fall 2015 Professor Brandon A. Jones Lecture 14: Probability and Statistics.
University of Colorado Boulder ASEN 5070: Statistical Orbit Determination I Fall 2015 Professor Brandon A. Jones Lecture 6: Linearization of OD Problem.
Curve Fitting Introduction Least-Squares Regression Linear Regression Polynomial Regression Multiple Linear Regression Today’s class Numerical Methods.
Engineering Analysis ENG 3420 Fall 2009 Dan C. Marinescu Office: HEC 439 B Office hours: Tu-Th 11:00-12:00.
University of Colorado Boulder ASEN 5070: Statistical Orbit Determination I Fall 2014 Professor Brandon A. Jones Lecture 30: Lecture Quiz, Project Details,
Colorado Center for Astrodynamics Research The University of Colorado 1 STATISTICAL ORBIT DETERMINATION EKF and Observability ASEN 5070 LECTURE 23 10/21/09.
University of Colorado Boulder ASEN 5070: Statistical Orbit Determination I Fall 2015 Professor Brandon A. Jones Lecture 17: Minimum Variance Estimator.
University of Colorado Boulder ASEN 5070: Statistical Orbit Determination I Fall 2015 Professor Brandon A. Jones Lecture 32: Gauss-Markov Processes and.
University of Colorado Boulder ASEN 5070 Statistical Orbit determination I Fall 2012 Professor George H. Born Professor Jeffrey S. Parker Lecture 9: Least.
Colorado Center for Astrodynamics Research The University of Colorado 1 STATISTICAL ORBIT DETERMINATION Kalman Filter with Process Noise Gauss- Markov.
University of Colorado Boulder ASEN 5070: Statistical Orbit Determination I Fall 2015 Professor Brandon A. Jones Lecture 22: Further Discussions of the.
University of Colorado Boulder ASEN 5070: Statistical Orbit Determination I Fall 2014 Professor Brandon A. Jones Lecture 10: Weighted LS and A Priori.
University of Colorado Boulder ASEN 5070 Statistical Orbit determination I Fall 2012 Professor George H. Born Professor Jeffrey S. Parker Lecture 10: Batch.
R. Kass/W03 P416 Lecture 5 l Suppose we are trying to measure the true value of some quantity (x T ). u We make repeated measurements of this quantity.
University of Colorado Boulder ASEN 5070: Statistical Orbit Determination I Fall 2015 Professor Brandon A. Jones Lecture 26: Cholesky and Singular Value.
University of Colorado Boulder ASEN 5070: Statistical Orbit Determination I Fall 2015 Professor Brandon A. Jones Lecture 15: Statistical Least Squares.
Statistics 350 Lecture 2. Today Last Day: Section Today: Section 1.6 Homework #1: Chapter 1 Problems (page 33-38): 2, 5, 6, 7, 22, 26, 33, 34,
Colorado Center for Astrodynamics Research The University of Colorado 1 STATISTICAL ORBIT DETERMINATION Statistical Interpretation of Least Squares ASEN.
ASEN 5070: Statistical Orbit Determination I Fall 2014
Simple Linear Regression
STATISTICAL ORBIT DETERMINATION Kalman (sequential) filter
ASEN 5070: Statistical Orbit Determination I Fall 2014
ASEN 5070: Statistical Orbit Determination I Fall 2015
Kin 304 Regression Linear Regression Least Sum of Squares
Ch3: Model Building through Regression
ASEN 5070: Statistical Orbit Determination I Fall 2015
ASEN 5070: Statistical Orbit Determination I Fall 2014
ASEN 5070: Statistical Orbit Determination I Fall 2015
BPK 304W Regression Linear Regression Least Sum of Squares
BPK 304W Correlation.
Regression Statistics
Consider Covariance Analysis Example 6.9, Spring-Mass
Presentation transcript:

University of Colorado Boulder ASEN 5070: Statistical Orbit Determination I Fall 2015 Professor Brandon A. Jones Lecture 19: Examples with the Batch Processor

University of Colorado Boulder  Exam 1 – Friday, October 9 ◦ Any exam related questions?  My office hours today in CCAR Meeting room instead of ECNT 420 2

University of Colorado Boulder  Everyone did well on these two quizzes  All answers are included in the slides as an appendix, but we will only go over two questions from Quiz 4 3

University of Colorado Boulder  Percent Correct: D2L Error  Consider the observation-state equation: In the case of a nonlinear estimation problem, which of the following are true: ◦ The observation-state relationship is linear with respect to deviation vectors ◦ The observation-state relationship is linear with respect to the nonlinear estimated state X ◦ We require an a priori x (deviation vector) to estimate the state using least squares ◦ We require an a priori X (nonlinear state) to estimate the state using least squares 4

University of Colorado Boulder  Percent Correct: 30%  In the case of nonlinear estimation using the linear batch filter, we attempt to estimate a state deviation vector x by solving for the vector that minimizes the sum of the observation residuals. By adding the state deviation vector to our best guess for the initial trajectory (X*), we get an updated state. To get this estimated state deviation vector, we require the observation deviation vector y. To solve for this, we use the predicted measurement G(X*,t). ◦ True ◦ False 5

University of Colorado Boulder 6 Illustration – Object in Ballistic Trajectory

University of Colorado Boulder  A cannonball has been launched with some uncertainty on the initial trajectory. We wish to: ◦ Estimate the initial state of the cannonball for future calibrations ◦ Determine where the cannonball went  We have some observations near the peak of the trajectory. 7

University of Colorado Boulder 8 Start of measurements  Object in ballistic trajectory under the influence of gravity Start of filter

University of Colorado Boulder  Object in ballistic trajectory under the influence of gravity  Equations of motion: 9  EOMs: Linear or Nonlinear?

University of Colorado Boulder  Object in ballistic trajectory under the influence of gravity  Observation Equations: 10  Obs. Eqns: Linear or Nonlinear?  Filter: Linear or Nonlinear?

University of Colorado Boulder  Filter: Linear or Nonlinear?  What do we need to solve via least squares? 11

University of Colorado Boulder  How do we get the STM for this problem? 12

University of Colorado Boulder  How do we get H_tilde for this problem?  What is H ? 13

University of Colorado Boulder  We have initial uncertainties on the a priori and the observations: 14

University of Colorado Boulder 15 Pierson Correlation Coeffs

University of Colorado Boulder 16

University of Colorado Boulder 17  Filter error smaller than measurement errors  Why does the uncertainty decrease and then increase?

University of Colorado Boulder  Illustrates error and 3σ bounds for data fit and prediction 18 Filter Span Predicted

University of Colorado Boulder  Should examine both the pre- and post-fit residuals: 19

University of Colorado Boulder 20

University of Colorado Boulder 21

University of Colorado Boulder  Observation Equations: 22 Station 1 Station 2

University of Colorado Boulder  How do we get H_tilde for this problem? 23

University of Colorado Boulder 24  Filter error smaller than measurement errors  Uncertainty decreases and then increases

University of Colorado Boulder  Illustrates error 3σ bounds for data fit and prediction 25

University of Colorado Boulder 26

University of Colorado Boulder 27

University of Colorado Boulder 28

University of Colorado Boulder 29

University of Colorado Boulder 30

University of Colorado Boulder 31

University of Colorado Boulder 32 Appendix: Lecture Quizzes

University of Colorado Boulder 33 Lecture Quiz 4

University of Colorado Boulder  Percent Correct: D2L Error  Consider the observation-state equation: In the case of a nonlinear estimation problem, which of the following are true: ◦ The observation-state relationship is linear with respect to deviation vectors ◦ The observation-state relationship is linear with respect to the nonlinear estimated state X ◦ We require an a priori x (deviation vector) to estimate the state using least squares ◦ We require an a priori X (nonlinear state) to estimate the state using least squares 34 89% 25% 20%

University of Colorado Boulder  Percent Correct: 30%  In the case of nonlinear estimation using the linear batch filter, we attempt to estimate a state deviation vector x by solving for the vector that minimizes the sum of the observation residuals. By adding the state deviation vector to our best guess for the initial trajectory (X*), we get an updated state. To get this estimated state deviation vector, we require the observation deviation vector y. To solve for this, we use the predicted measurement G(X*,t). ◦ True ◦ False 35

University of Colorado Boulder  Percent Correct: 93%  Consider the weighted least-squares cost function J(x). We have two observation errors e 1 and e 2. The weights for those observations are w 1 =3 and w 2 =2. Which of the following provides the best solution? ◦ e 1 =1, e 2 =1 ◦ e 1 =2, e 2 =1 ◦ e 1 =1, e 2 =2 ◦ e 1 =1, e 2 =1/2 36

University of Colorado Boulder  Percent Correct: 84%  In the weighted least-squares estimator, the H matrix no longer needs to be full rank. ◦ True ◦ False 37

University of Colorado Boulder  Percent Correct: 91%  For a linear estimation problem solved via the batch filter, we require a priori information to obtain a solution. ◦ True ◦ False 38

University of Colorado Boulder 39 Lecture Quiz 5

University of Colorado Boulder  Percent Correct: 95%  The inverse of the variance-covariance matrix is symmetric ◦ True ◦ False 40

University of Colorado Boulder  Percent Correct: 76%  Which of the following lists of numbers has the largest variance? A: [ 0.0, 0.5, 1.0 ] B: [ 0.0, 0.25, 0.5, 0.75, 1.0 ] C: [ 0.0, 0.1, 0.2, 0.3, …, 0.8, 0.9, 1.0 ] ◦ A ◦ B ◦ C ◦ They all have the same variance 41

University of Colorado Boulder  Percent Correct: 86%  If X and Y are independent random variables drawn from the standard normal distribution and Z = X+Y, which of the following best describes the probability density of Z? ◦ U(0,1) (uniform distribution with range [0,1]) ◦ U(0,2) (uniform distribution with range [0,2]) ◦ A normal distribution with mean 0.0 ◦ A normal distribution with mean

University of Colorado Boulder  Percent Correct: 90%  If X is the number of people who fall asleep during an average ASEN 5070 lecture and “X” is drawn from U(0,2), then what is the expected value for the total number of instances of people falling asleep after 42 independent lectures? ◦ 0 ◦ 16 ◦ 30 ◦ Incorrect, but brownie points awarded!

University of Colorado Boulder  Percent Correct: 88% ◦ Let f(x) be a probability density function. Which of the following are true? 44 90% 98% 100% 2%