Next Page Previous Page Home Tools & Aids Search Handbook

4.   Process Modeling - Detailed Table of Contents  [4.]


The goal for this chapter is to present the background and specific analysis techniques needed to construct a statistical model that describes a particular scientific or engineering process. The types of models discussed in this chapter are limited to those based on an explicit mathematical function. These types of models can be used for prediction of process outputs, for calibration, or for process optimization.
  1. Introduction to Process Modeling  [4.1.]
    1. What is process modeling?  [4.1.1.]
    2. What terminology do statisticians use to describe process models?  [4.1.2.]
    3. What are process models used for?  [4.1.3.]
      1. Estimation  [4.1.3.1.]
      2. Prediction  [4.1.3.2.]
      3. Calibration  [4.1.3.3.]
      4. Optimization  [4.1.3.4.]
    4. What are some of the different statistical methods for model building?  [4.1.4.]
      1. Linear Least Squares Regression  [4.1.4.1.]
      2. Nonlinear Least Squares Regression  [4.1.4.2.]
      3. Weighted Least Squares Regression  [4.1.4.3.]
      4. LOESS (aka LOWESS)  [4.1.4.4.]

  2. Underlying Assumptions for Process Modeling  [4.2.]
    1. What are the typical underlying assumptions in process modeling?  [4.2.1.]
      1. The process is a statistical process.  [4.2.1.1.]
      2. The means of the random errors are zero.  [4.2.1.2.]
      3. The random errors have a constant standard deviation.  [4.2.1.3.]
      4. The random errors follow a normal distribution.  [4.2.1.4.]
      5. The data are randomly sampled from the process.  [4.2.1.5.]
      6. The explanatory variables are observed without error.  [4.2.1.6.]

  3. Data Collection for Process Modeling  [4.3.]
    1. What is design of experiments (DOE)?  [4.3.1.]
    2. Why is experimental design important for process modeling?  [4.3.2.]
    3. What are some general design principles for process modeling?  [4.3.3.]
    4. I've heard some people refer to "optimal" designs, shouldn't I use those?  [4.3.4.]
    5. How can I tell if a particular experimental design is good for my application?  [4.3.5.]

  4. Data Analysis for Process Modeling  [4.4.]
    1. What are the basic steps for developing an effective process model?  [4.4.1.]
    2. How do I select a function to describe my process?  [4.4.2.]
      1. Incorporating Scientific Knowledge into Function Selection  [4.4.2.1.]
      2. Using the Data to Select an Appropriate Function  [4.4.2.2.]
      3. Using Methods that Do Not Require Function Specification  [4.4.2.3.]
    3. How are estimates of the unknown parameters obtained?  [4.4.3.]
      1. Least Squares  [4.4.3.1.]
      2. Weighted Least Squares  [4.4.3.2.]
    4. How can I tell if a model fits my data?  [4.4.4.]
      1. How can I assess the sufficiency of the functional part of the model?  [4.4.4.1.]
      2. How can I detect non-constant variation across the data?  [4.4.4.2.]
      3. How can I tell if there was drift in the measurement process?  [4.4.4.3.]
      4. How can I assess whether the random errors are independent from one to the next?  [4.4.4.4.]
      5. How can I test whether or not the random errors are distributed normally?  [4.4.4.5.]
      6. How can I test whether any significant terms are missing or misspecified in the functional part of the model?  [4.4.4.6.]
      7. How can I test whether all of the terms in the functional part of the model are necessary?  [4.4.4.7.]
    5. If my current model does not fit the data well, how can I improve it?  [4.4.5.]
      1. Updating the Function Based on Residual Plots  [4.4.5.1.]
      2. Accounting for Non-Constant Variation Across the Data  [4.4.5.2.]
      3. Accounting for Errors with a Non-Normal Distribution  [4.4.5.3.]

  5. Use and Interpretation of Process Models  [4.5.]
    1. What types of predictions can I make using the model?  [4.5.1.]
      1. How do I estimate the average response for a particular set of predictor variable values?  [4.5.1.1.]
      2. How can I predict the value and and estimate the uncertainty of a single response?  [4.5.1.2.]
    2. How can I use my process model for calibration?  [4.5.2.]
      1. Single-Use Calibration Intervals  [4.5.2.1.]
    3. How can I optimize my process using the process model?  [4.5.3.]

  6. Case Studies in Process Modeling  [4.6.]
    1. Load Cell Calibration  [4.6.1.]
      1. Background & Data  [4.6.1.1.]
      2. Selection of Initial Model  [4.6.1.2.]
      3. Model Fitting - Initial Model  [4.6.1.3.]
      4. Graphical Residual Analysis - Initial Model  [4.6.1.4.]
      5. Interpretation of Numerical Output - Initial Model  [4.6.1.5.]
      6. Model Refinement  [4.6.1.6.]
      7. Model Fitting - Model #2  [4.6.1.7.]
      8. Graphical Residual Analysis - Model #2  [4.6.1.8.]
      9. Interpretation of Numerical Output - Model #2  [4.6.1.9.]
      10. Use of the Model for Calibration  [4.6.1.10.]
      11. Work This Example Yourself  [4.6.1.11.]
    2. Alaska Pipeline  [4.6.2.]
      1. Background and Data  [4.6.2.1.]
      2. Check for Batch Effect  [4.6.2.2.]
      3. Initial Linear Fit  [4.6.2.3.]
      4. Transformations to Improve Fit and Equalize Variances  [4.6.2.4.]
      5. Weighting to Improve Fit  [4.6.2.5.]
      6. Compare the Fits  [4.6.2.6.]
      7. Work This Example Yourself  [4.6.2.7.]
    3. Ultrasonic Reference Block Study  [4.6.3.]
      1. Background and Data  [4.6.3.1.]
      2. Initial Non-Linear Fit  [4.6.3.2.]
      3. Transformations to Improve Fit  [4.6.3.3.]
      4. Weighting to Improve Fit  [4.6.3.4.]
      5. Compare the Fits  [4.6.3.5.]
      6. Work This Example Yourself  [4.6.3.6.]
    4. Thermal Expansion of Copper Case Study  [4.6.4.]
      1. Background and Data  [4.6.4.1.]
      2. Rational Function Models  [4.6.4.2.]
      3. Initial Plot of Data  [4.6.4.3.]
      4. Quadratic/Quadratic Rational Function Model  [4.6.4.4.]
      5. Cubic/Cubic Rational Function Model  [4.6.4.5.]
      6. Work This Example Yourself  [4.6.4.6.]

  7. References For Chapter 4: Process Modeling  [4.7.]

  8. Some Useful Functions for Process Modeling  [4.8.]
    1. Univariate Functions  [4.8.1.]
      1. Polynomial Functions  [4.8.1.1.]
        1. Straight Line  [4.8.1.1.1.]
        2. Quadratic Polynomial  [4.8.1.1.2.]
        3. Cubic Polynomial  [4.8.1.1.3.]
      2. Rational Functions  [4.8.1.2.]
        1. Constant / Linear Rational Function  [4.8.1.2.1.]
        2. Linear / Linear Rational Function  [4.8.1.2.2.]
        3. Linear / Quadratic Rational Function  [4.8.1.2.3.]
        4. Quadratic / Linear Rational Function  [4.8.1.2.4.]
        5. Quadratic / Quadratic Rational Function  [4.8.1.2.5.]
        6. Cubic / Linear Rational Function  [4.8.1.2.6.]
        7. Cubic / Quadratic Rational Function  [4.8.1.2.7.]
        8. Linear / Cubic Rational Function  [4.8.1.2.8.]
        9. Quadratic / Cubic Rational Function  [4.8.1.2.9.]
        10. Cubic / Cubic Rational Function  [4.8.1.2.10.]
        11. Determining m and n for Rational Function Models  [4.8.1.2.11.]
Home Tools & Aids Search Handbook Previous Page Next Page