Compare Calibrations to Data

Setting Up the Data Set

You can use the Data Sets view in CAGE to compare features, tables, and models with experimental data. You can use data sets to plot the features, tables, etc., as tabular values or as plots on a graph.

Data sets enable you to view the data at a set of operating points. You can determine the set of operating points yourself, using Build Grid. Alternatively, you can import a set of experimental data taken at a series of operating points. These operating points are not the same as the breakpoints of your tables.

This tutorial takes you through the basic steps required to compare a completed feature calibration to a set of experimental data.

Start CAGE by typing

cage

To set up the data set tutorial, you need to

Open an existing calibration.
Import the experimental data.
Add the Torque feature to the data set.

Your data set contains all the input factors and output factors required. As the imported data contains various operating points, this information is also included in the data set.

Opening an Existing Calibration

For this tutorial, use the file datasettut.cag, found in the matlab\toolbox\mbc\mbctraining directory.

To open this file,

Select File > Open Project.
In the file browser, select datasettut.cag and click Open.
This opens a file that contains a complete calibrated feature with its associated models and variables. This particular feature is a torque calibration, using a torque table (labeled T1) and modifiers for spark (labeled T2) and air/fuel ratio (labeled T3).
Select File > New > Data Set to add a new data set to your session.

This automatically switches you to the Factor Information pane of the data set display.

Importing Experimental Data into a Data Set

To import data into a data set,

Select File > Import > Data > File.
In the file browser, select meas_tq_data.xlsx from the mbctraining directory, and click Open.
This set of data includes six columns of data, the test cell settings for engine speed (RPM), and the measured values of torque (tqmeas), engine speed (nmeas), air/fuel ratio (afrmeas), spark angle (spkmeas), and load (loadmeas).
The Data Set Import Wizard asks which of the columns of data you would like to import. Click Next to import them all.
The following screen asks you to associate variables in your project with data columns in the data.
Highlight afr in the Project Assignments column and afrmeas in the Data Column, then click the assign button, shown.
Repeat this to associate load with loadmeas, n with RPM, and spk with spkmeas. The dialog box should be the same as shown.
Click Finish to close the dialog box.
Note
If you need to reassign any inputs after closing this dialog box you can click in the toolbar or select Data > Assign.

Adding an Item to a Data Set

To add the Torque feature to the data set,

Highlight the Torque feature in the lower list of Project Expressions.
Select Data > Factors > Add to Data Set.

This adds two objects to the data set: Torque: Model and Torque: Strategy. These two objects make up the Torque feature.

Torque: Model is the model used as a reference point to calibrate the feature.
Torque: Strategy is the values of the feature at these operating points.

When these steps are complete, the list of factors includes four input factors and four output factors, as shown.

Cage browser dialog box

Comparing the Items in a Data Set

Viewing the Data Set as a Table

By viewing the data set, you can compare experimental data with calibrations or models in your project.

Click in the toolbar to view the data set as a table of values.

Table of data containing 19 rows and n, load, aft, spk, meas, tqmeas, Torque model, and Torque strategy columns

In the table, the input cells are white and the output cells are gray. Select the Torque: Strategy column header to see the view shown. The selected column turns blue and the column headers of the strategy's inputs (n, load, afr and spk) turn cream. Column headers are always highlighted in this way when they are associated with the currently selected column (such as model inputs, strategy inputs or linked columns).

In addition to viewing the columns, you can use data sets to create a column that shows the difference between two columns:

Select the tqmeas and Torque: Strategy columns by using Ctrl+click.
Select Create Error from the right-click menu on either column header.

This creates another column that is the difference between tqmeas and Torque: Strategy. Note that all the columns that are inputs to this new column have highlighted headers.

Table of data containing 19 rows and n, load, aft, spk, nmeas, tqmeas, Torque model, and Torque strategy columns

The error column is simply the difference between tqmeas and Torque: Strategy. This provides a simple way of comparing the feature and the measured data.

Viewing the Data Set as a Plot

Click or select View > Plot to view the data set as a plot.
The lower pane lists all the output expressions in the data set and in the project.
Use Ctrl+click to select tqmeas and Torque: Strategy from the lower list.
Change the x-axis factor to n from the drop-down menu.
This displays the calibrated values of torque from the feature, and the measured values of torque from the experimental data, against the test cell settings for engine speed.
Clearly there is some discrepancy between the two.

Displaying the Error

View the error between the calibrated and measured values of torque.

Cage browser dialog box

Select tqmeas_minus_Torque from the lower list (Output Expressions).
For the y-axis factor, select Absolute Relative Error (tqmeas - Torque) from the drop-down menu.

As you can see, there seems to be no particular correlation between engine speed and the error in the calibration.

Coloring the Display

Select Color by Value from the right-click menu on the graph.
From the Color by drop-down menu, select load.

In this display, you can see that some of the low values of load display a high error.

Limiting the Range of the Colors

Color range corresponding to load

To view the colors in more detail, you can limit the range of the colors:

Select the Limit range box (or you could right-click the graph and select Restrict Color to Limits).
Set the minimum value of the color range to be as low as possible by dragging the minimum value down.
Set the maximum value of the color range to be around 0.4.

As the low values of load are causing large errors, it would be wise to reexamine the calibration, particularly at small values of load.

Reassigning Variables

You can alter the data set by changing which variables are used for project expressions.

Instead of using the test cell settings for the engine speed (RPM), you might want to use the measured values of engine speed (nmeas). So you have to reassign the variable n to nmeas.

To reassign n,

Click or select Data > Assign.
In the dialog box that appears, select n from the Project Assignments pane and nmeas from the Data Columns pane.
Click the assign button.