Model Browser User's Guide

Getting Started

What Is the Model-Based Calibration Toolbox?

About the Model Browser

About CAGE

How to Use This Manual

System Requirements

Hardware Requirements

Operating System Requirements

Required MathWorks Products

Optional MathWorks Products

Quickstart Tutorial

Two-Stage Models

Starting the Toolbox

Setting Up the Model

Setting Up the Local Model

Setting Up the Global Model

Selecting Data

Specifying the Response Model

Verifying the Model

Verifying the Local Model

Verifying the Global Model

Selecting the Two-Stage Model

Comparing the Local Model and the Two-Stage Model

Maximum Likelihood Estimation

Response Node

Exporting the Model

Design of Experiment Tutorial

What Is Design of Experiment?

Design Styles

Starting the Design Editor

Optimal Designs

Start Point Tab

Candidate Set Tab

Algorithm Tab

Design Displays

Prediction Error Variance Viewer

Improving the Design

Classical Designs

Classical Design Browser

Set Up and View a Classical Design

Design Evaluation Tool

Improving the Design

Space-Filling Designs

Setting Up a Space-Filling Design

Applying Constraints

Saving Designs

Data Editor Tutorial

Loading the Data

Entering the Data Editor

Loading a Data File

Viewing and Editing the Data

User-Defined Variables & Filtering

Defining a New Variable

Applying a Filter

Sequence of Variables

Deleting and Editing Variables and Filters

Storage

Test Groupings

GUI Reference

Project Level: Startup View

Project Level: Toolbar

Project Level: Menus

Model Tree

Test Plans

Creating a New Test Plan

Creating New Test Plan Templates

Test Plan Level

Block Diagram Representations of Models

Test Plan Level: Toolbar and Menus

Setting Up Models

Instant One-Stage Model Setup

Instant Two-Stage Model Setup

Setting Up Inputs

Global Model Setup

Global Linear Models: Polynomials and Hybrid Splines

Global Model Class: Radial Basis Function

Global Model Class: Hybrid RBF

Global Model Class: Multiple Linear Models

Global Model Class: Free Knot Spline

Local Model Setup

Local Model Class: Polynomials and Polynomial Splines

Local Model Class: Truncated Power Series

Local Model Class: Free Knot Spline

Local Model Class: Growth Models

Local Model Class: Linear Models

Local Model Class: User-Defined Models

Local Model Class: Transient Models

Covariance Modeling

Correlation Models

Transforms

Designing Experiments

Data

The Data Editor

Data Loading and Merging

User-Defined Variables and Filtering

Variable and Filter Editors

Storage

Test Groupings

Data Wizard

Data Selection Window

New Response Models

Datum Models

Local Level

Local Special Plots

Local Scatter Plots

Diagnostic Statistics Pane

Response Features List Pane

Test Notes Pane

Local Level: Toolbar

Local Level: Menus

Model Menu (Local Level)

View Menu (Local Level)

Outliers Menu (Local Level)

Outlier Selection Criteria

Data Tab

Global Level

Global Special Plots

Global Scatter Plots

Global Level: Toolbar

Global Level: Model-Specific Tools

Global Level: Menus

Selecting Models

Select

Model Selection Window

Tests View

Predicted/Observed View

Response Surface View

Likelihood View

RMSE View

Residuals View

Cross Section View

MLE

MLE Settings

Response Level

Response Level: Toolbar and Menus

Model Evaluation Window

Model Evaluation Using Other Data

Exporting Models

What Is Exported?

Evaluating Models in the Workspace

The Design Editor

Design Styles

Design Editor Displays

Creating a Classical Design

Creating a Space Filling Design

Setting Up a Space Filling Design

Creating an Optimal Design

Optimal Design: Start Point Tab

Optimal Design: Candidate Set Tab

Optimal Design: Algorithm tab

Adding Design Points

Fixing, Deleting, and Sorting Design Points

Saving and Importing Designs

Applying Constraints

Constraint Types

Importing Constraints

Prediction Error Variance Viewer

Technical Documents

Linear Regression

Stepwise Regression

Definitions

Prediction Error Variance

High-Level Model Building Process Overview

Univariate Model Building Process Overview

Stepwise Regression Techniques

Stepwise Table

Box-Cox Transformation

Linear Model Statistics Displays

Summary Statistics

ANOVA Table

Diagnostic Statistics

PRESS Statistic

Pooled Statistics

Design Evaluation Tool

Design Matrix

Full FX Matrix

Model Terms

Z2 Matrix

Alias Matrix

Z2.1 Matrix

Regression Matrix

Coefficient Information

Standard Error

Hat Matrix

|X'X|

Raw Residual Statistics

Degrees of Freedom Table

Design Evaluation Graphical Displays

Export of Design Evaluation Information

Maximum Likelihood Estimation

Two-Stage Models for Engines

Local Models

Global Models

Two-Stage Models

Prediction Error Variance for Two-Stage Models

Global Model Selection

References

Local Model Definitions

Local Models and Associated Response Features

Polynomial Models

Polynomial Splines

Truncated Power Series Basis (TPSBS) Splines

Free Knot Splines

Three Parameter Logistic Model

Morgan-Mercer-Flodin Model

Four-Parameter Logistic Curve

Richards Curves

Weibul Growth Curve

Exponential Growth Curve

Gompertz Growth Model

Neural Networks

User-Defined Models

To Begin

Template File

Subfunctions

Optional Subfunctions

Checking into MBC

Check In

MBC Modeling

Transient Models

The Simulink Model

Parameters

Checking the Model

Transient Models

Template File

Subfunctions

Checking into MBC

Check In

MBC Modeling

Data Loading Application Programming Interface

Data Function Prototype

Input Arguments

Output Arguments

Data Function Check In

Radial Basis Functions

Guide to Radial Basis Functions for Model Building

Types of Radial Basis Functions

Gaussian

Thin-Plate Spline

Logistic Basis Function

Wendland's Compactly Supported Function

Multiquadrics

Reciprocal Multiquadrics

Fitting Routines

Center Selection Algorithms

Rols

RedErr

WiggleCenters

CenterExchange

Lambda Selection Algorithms

IterateRidge

IterateRols

StepItRols

Width Selection Algorithms

TrialWidths

WidPerDim

Prune Functionality

Statistics

GCV Criterion

GCV for Ridge Regression

GCV for Rols

References

Hybrid Radial Basis Functions

Width Selection Algorithm: TrialWidths

Lambda and Term Selection Algorithms: Interlace

Lambda and Term Selection Algorithms: TwoStep

Tips for Modeling with Radial Basis Functions

Plan of Attack

How Many RBFs to Use

Width Selection Algorithms

Which RBF to Use

Lambda Selection Algorithms

Center Selection Algorithms

General Parameter Fine-Tuning

Hybrid RBFs

Index

Getting Started