Figure 1: Methods for extracting segments of data available in Simcenter Testlab Neo Process Designer.
All of these methods output a new file containing only a extracted segment of the original file. Additional methods (spectrums, overall level, etc) can be connected in turn to analyze any extracted segments.
The following examples for segmenting data are covered in this article: 1. “Input" Method 1.1 Interactive Selection of Segments 1.2 Selecting Different Segments Per Run 2. “Extract Segments (Rule-Based)” Method 2.1 Using the Method 2.2 Two Conditions: Creating an Index Channel 2.3 Recording Events or Ratings 3. “Split by Position” Method 3.1 Export GPS Measurement to Google Earth 3.2 Create Triggers in Google Earth 3.3 Export Trigger KML File from Google Earth 3.4 Use Triggers KML File with "Split by Position" in Simcenter Testlab Neo 4. “Decompose” Method 5. “Extract Segment (Block)” Method 6. Results
The "Input" method can be used to select segments of data versus time.
1.1 Interactive Selection of Segments
Segments of data can be highlighted manually in a display and then used with the “Input” method.
After displaying a time history, hold down the CTRL and ALT keys at the same time while simultaneously dragging from left to right to highlight segments as shown in Figure 2.
Figure 2: To select segments, hold down the CTRL and ALT keys at the same time and simultaneously dragging the mouse from left to right.
Multiple segments can be selected this way. If only one segment is to be selected, use only the CTRL key without the ALT key.
With the segments selected, click on the "Input" method in the process. A “Overwrite Segments” dialog should appear as shown in Figure 3 below.
Figure 3: Click on the "Input" method. A "Overwrite Segments" message should appear.
The "Overwrite Segments" message typically appears the first time only. If the segments in the display are changed again, still click on the "Input" method. No message will appear, but the segments will update.
To verify that the segments were selected or updated, double click on the "Input" method to view the Properties as shown in Figure 4.
Figure 4. Double click on the "Input" method to verify the segments were selected. Click on the “…” button next to "Segments" to get a list of time segments to be saved as runs.
In the "Input" method "Properties" dialog:
Change the "Input Data" field to "Save in Run". This will create separate runs for each selected segment.
Cick on the button “…” next to “Segments”. The selected segments should appear in a list.
It is possible to give unique names to each of the segments selected in the display. In the “Segment List” dialog in the “Run suffix” field. After processing the data, this is the name that will be given to the run.
When finished, press OK in the “Segment List” menu and press “Close” on the Input method property dialog.
After adding the data to the "Input Basket", press the "Run" button to process the data. New runs will be created with extracted time data. If additional methods (like spectral map, orders, etc) were added, the processed data will also be stored in the runs.
1.2 Selecting Different Segments Per Run
It is also possible to select different segments per run/recording of data. This might be helpful if an event happens at a different time in each recording.
To do this, double click on the input method, and switch the "Segment Mode" from "All Runs" to "Per Run" as shown in Figure 5:
Figure 5: To select a different time segment for different recordings, double click on the "Input" method and change "Segment Mode" from "All Runs" to "Per Run".
Then click on the "Segments" button with the "..." to set different segments per run as shown in Figure 6:
Figure 6: In the "Input" method properties, after selecting "Per Run", click on the segments area to define different time intervals for each run.
In the "Segment List" menu, the first column contains pulldowns to select the runs. A different segment can be set per run.
The “Extract Segments (Rule-Based)” method can be used to identify segments of data to process based on the values contained in a data channel. This allows the segment to be identified automatically by the software without manual user intervention. For example, a segment could be extracted based on certain torque or rpm or speed values.
As an example, in the data shown in Figure 7 below, it is desired to calculate a spectrum at a speed of 65 kilometers per hour.
Figure 7: Top – Varying speed trace with one segment of constant speed. Bottom – Corresponding data trace to be processed.
In the recorded data, the speed is not constant. Rather than trimming and cutting the data manually, the “Extract Segments (Rule-Based)” method can be used to automatically identify and process data at a constant speed of 65 kilometers per hour.
To make the method available in the Method Library, go to "Files -> Add-ins" and turn on “Trackside Validation” add-in (17 tokens). The “Extract Segments (Rule-Based)” will be available as shown in Figure 8.
Figure 8: The “Extract Segments (Rule-Based)” will be available after turning on the “Track-side Validation” add-in.
Place the “Extract segments (rule-based)” method between the “Input” method that contains the data and the “Spectrum average” processing method.
Double click on the “Extract segments (rule-based)” method to edit the Properties (Figure 9).
Figure 9: Properties dialog for “Extract segments (rule-based)” method.
In the “Detection Mode” pulldown, select either “First Occurrence” or “All Occurrences”:
First Occurrence: Once the criteria set in “Segment Selection” are met, a single segment will be extracted from the data trace.
All Occurrences: Multiple segments can be selected from the data if there is more than one instance where the “Segment Selection” criteria are met in the recording. When selecting this option, there are further options to "Join" together the extracted segments into a single run, or keep the extracted segments "Separate".
Click on the “…” button next to the “Segment selection” to create rules for extracting the data as shown in Figure 10.
Figure 10: Rule definition in the properties of “Extract segments (rule-based)” method.
In the “Extract segments (rule-based)” method, a reference signal needs to be selected. The reference signal is a single data channel to be used as the basis for identifying segments. If a combination of conditions on different channels is desired to identify the segment, use the "Calculator" method as described in the next section.
The rest of the rule definition menu has three main options for identifying segments. All values must include the Engineering Units to work.
Begin or End: Define a beginning and end condition for the segment.
Extract: Identify data that falls either inside or outside two different values.
Triggers: A trigger level can be defined. For example, if the specified value exceeds a certain threshold, the preceding 5 seconds of data can be captured.
In the screenshot above, the “Extract” option was used to identify segments where the speed channel was between 64 and 66 kilometers per hour (km/h).
The engineering units for the selected reference channel must be specified in the value field (Example: km/h, Nm, rpm). To avoid confusion, spell the units exactly as shown in the Y-axis legend of the display. The field is case-sensitive.
With the rules defined, click on the “Run” button in the lower left corner to process the data. The result of the “Extract segments (rule-based)” method is shown in Figure 11 below.
Figure 11: After running the process with the “Extract segments (rule-based)” method a new segment trace is extracted from the original recording. Top graph shows the original speed channel (blue) used to extract a segment (green) with speeds that fell between 64 and 66 kilometers per hour (kph). The bottom graph is the corresponding original data channels (red and green) and the extracted segments (blue and cyan).
Both the original time signals and extracted segments can be overlaid and viewed. This can be done by clicking on the “Active Analysis”. If the extracted segment is satisfactory, press “Accept”.
Suppose two conditions are to be met to extract a segment. In this case, a “Calculate” method (for time data) can be used in conjunction with the “Extract segments (rule-based)” method.
In the process shown in Figure 12, an “Index” channel is created which has a value of 0 by default, but will have a value of 1 when 3500 rpm and 100 Nm of torque are both exceeded:
Figure 12: Calculate method to create an Index channel
In the formula set (use the “…” button in the Properties dialog of “Calculate” to set):
Torque/moment is assigned to R1 while rotational speed/rpm is assigned to R2.
R2 has a threshold of 3500 rpm (the units must be specified within square brackets) and assigned to V1.
R1 has a threshold of 100 Nm (the units must be specified within square brackets) and assigned to V2.
V3 is a logical AND operation between V1 and V2. The Index channel output is a zero unless both V1 and V2 are true. When both are true, the Index channel will have a value of 1.
Example inputs and output of the “Calculate” method are shown in Figure 13.
Figure 13: Torque moment input (red) and rotational speed (green) inputs to the calculate method. When torque moment exceeds 100 Nm and rotational speed exceeds 3500 rpm, the output Index channel (blue) will have a value of 1.
The Index channel takes into account both conditions and only has values of zero or one (when both conditions are true). The Index channel can be fed to the “Extract segments (rule-based)” method as shown in Figure 14:
Figure 14: The “Extract segments (rule-based)” method settings to extract based on a logical index channel that alternates between numerical values of 0 and 1.
The segment is extracted when a threshold of 0.5 is exceeded. Notice that the start requires an upward slope while the end requires a downward slope.
2.3 Recording Events or Ratings
Simcenter Testlab Neo Time Data Acquisition has a marker and ratings feature (Figure 15) that can be used during acquisition.
Figure 15: The event marker and ratings feature of Time Data Acquisition can be used to help identify segments for automatic processing.
Numerical keystrokes from the keyboard of the acquisition PC are recorded along with the data. These could be ratings or event markers of the users choosing. This data can be used with the “Extract segments (rule-based)” method to segment the data.
The “Split by Position” method can bs used to identify segments of data to process using GPS/GNSS position on the earth. This could be used to identify specific road surfaces, etc.
To make the “Split by Position” method available in the Method Library of Simcenter Testlab Process Designer, go to “File -> Add-ins” and turn on “Trackside Validation” add-in (17 tokens). The “Split by Position” will be available as shown in Figure 16.
Figure 16: The “Split by Position” method can extract segments based on GPS/GNSS location. Turn on the “Track-side Validation” add-in to make it available in the Method Library.
Steps to use the "Split by Position" method:
Create triggers: Triggers (each trigger is a line defined by two points) need to be defined. Triggers can be created in Google Earth (either online version or downloaded version).
Names: The triggers need to have specific names that follow a certain convention. The name consists of three parts (Beginning, middle, end) separated by a colon symbol. In the last two parts only numbers (negative or positive) or the wild card * symbol are valid. For example “Name:*:1” or “Name:1:-1”.
Beginning: Name is a free field. This will be used as the extracted segment run name in Simcenter Testlab Neo after running the process with “Split by Position” method.
Middle:
If a sequence of triggers is desired, the middle number is the identifier number of the preceding trigger. Data will be recorded from the preceding trigger only.
Using the wildcard symbol * means no data from a previous trigger is needed to detect the current trigger. For example, this might be useful for a start trigger condition.
End: The last number is the identifier number assigned to the trigger. A positive number means the data after the trigger is to be stored. A negative number means data after the trigger is not to be stored. Any negative number can be used.
Export the triggers to KML (Keyhole Markup Language) format from Google Earth.
Import the KML trigger file to use with the “Split by Position” method in Simcenter Testlab Neo.
Run the process to extract the segments.
In the example below, a specific segment of road (yellow bracket) needs to be trimmed from a measurement (red line) as shown in Figure 17.
Figure 17: With a Google Earth plug-in in Simcenter Testlab Neo, the GPS location (red line) of the acquired data is shown. The yellow indicates the desired road segment to be analyzed. Right click in the map to switch from Satellite to Map view.
Clicking on either the Latitude or Longitude channel in the throughput file will show the route in Simcenter Testlab Neo.
Note that Simcenter Testlab Neo does not have the Google Earth display active by default (due to export restrictions). Contact your Siemens sales or support contact to request a license to add the capability at no cost if in an eligible country.
3.1 Export GPS Measurement to Google Earth
Right click on the throughput file in the navigation tree and choose ”Export…” as shown in Figure 18 below.
Figure 18: To create a KML file from the measurement, right click on the throughput file and choose “Export…”. When selecting KML as the export format, only GPS location data will be exported.
After selecting “Export…”, choose “KML” as the export file format. When exporting to KML only the GPS location data is exported. All other channels (acceleration, strain, etc) are not exported.
3.2 Create Triggers in Google Earth
Go to Google Earth (either web version at https://earth.google.com or the downloaded application) to define "triggers" around the beginning and end of the desired segment of road.
Create a new “Project” within Google Earth and import the KML file that was exported from Simcenter Testlab (Figure 19). Importing the original measurement will make it easier to select appropriate triggers in Google Earth.
Figure 19: Open a new project in Google earth and import the KML file (Open local KML file).
The GPS positions corresponding to the Simcenter Testlab recording are now shown in Google Earth (Figure 20):
Figure 20 The location of the vehicle during the measurement is shown in Google Earth.
Now click on the “Add path or polygon” icon. Define a start trigger by clicking on two points in Google Earth as shown in Figure 21.
Figure 21: Click on two points that cross the GPS path in Google Earth to define a single start trigger.
To be used by the “Split by GPS Postion” method in Simcenter Testlab Neo, the trigger needs a very specific name (see below for guidance). Repeat the operation again for the end trigger (Figure 22):
Figure 22: A start (light blue) and end trigger (light green) defined in Google Earth.
In Google Earth, give the start and end trigger specific names.
The start trigger is called “Euclid_Road_North:*:1”. It has a beginning, middle, and end separated by the “:” symbol:
Beginning: Euclid_Road_North is the name of the trigger and will be the name of the run corresponding to extracted data in Simcenter Testlab Neo.
Middle: The * symbol means the trigger will be activated by crossing it. The * is a like a wild card, the trigger will always be activated.
End: The number 1 is an identifier number to the trigger. Assigning a positive number means the data after the trigger is to be stored.
The end trigger is called “Euclid_Road_North:1:-1”
Beginning: Euclid_Road_North is the name of the trigger and will be the name of the run corresponding to extracted data in Simcenter Testlab Neo.
Middle: The number 1 means that data starting with trigger identifier 1 is to be saved. It only works if the start trigger with identifier 1 was the previous trigger that was activated. Any other triggers are ignored.
End: The number -1 means no data after this trigger is to be saved. The negative number effectively makes this the end trigger. Any negative number can be used in this manner, but for easier book keeping the number -1 was used.
3.3 Export Trigger KML File from Google Earth
After creating the triggers, but before exporting from the web based Google Earth to a KML file, be sure to delete the “GPS Points” and “GPS Track” as shown in Figure 23.
Figure 23: Delete both “Gps Points” and “Gps track” before exporting the triggers to a KML to be used in Simcenter Testlab Neo.
For Simcenter Testlab Neo, the KML file must only contain the triggers, not the “Points” and “Track”. If using the Google Earth application, deleting “Gps Points” and “Gps Track” is not needed because the triggers alone can be exported easily.
With the deletions completed, choose “File -> Export as KML file” in Google Earth to export the triggers as shown in Figure 24.
Figure 24: From Google Earth, choose “File -> Export as KML file” for use in the “Split by Position” method in Simcenter Testlab Neo Process Designer.
The newly exported trigger KML file is ready for use in Simcenter Testlab Neo Process Designer.
3.4 Use Triggers KML File with "Split by Position" in Simcenter Testlab Neo
With the trigger KML file created in Google Earth, go back to Simcenter Testlab Neo Process Designer. Double click on the “Split by position” method in Simcenter Testlab Process Designer to set the Properties.
In the Properties dialog, select the KML file in the “Track Definition” parameter (Figure 25):
Figure 25: In the Properties of the “Split by Position” method select the KML file under “Track Definition”.
Note that the setting “Export Track Definition” is only used to transfer the KML (perhaps the external file was lost) to another process or project.
Then click “Run” to execute the process. After accepting the result, the extracted segment will be stored in a run with the name of the trigger defined in Google Earth. The run will have a folder called “Process” that contains the extracted segment.
The “Decompose” method is part of the “Interactive Analysis” add-in. It can be used to create time histories based on an index channel.
For example, the gear (1st gear, 2nd gear, 3rd gear, etc) could act as an index channel (Figure 26) in the 900 second recording shown below:
Figure 26: Gear channel has six different values (zero thru five) that correspond to 1st thru 6th gear and a simultaneously measured temperature channel.
In Figure 27 below, the “Decompose” method is setup to create new data traces based on the six different gears:
Figure 27: The “Decompose” method will create six separate traces corresponding to each gear.
After running the method, six new traces are created from the temperature recording. Each recording contains the temperature data that corresponds to one gear as shown in Figure 28 below:
Figure 28: The original 900 second temperature recording is decomposed into six different recordings each corresponding to a specific gear.
As a result of the decompose, the temperature range increases with the gear. These new segments could be used to calculate the average or peak temperature for each gear.
Each trace is stitched together from data from different times in the original recording. The “Decompose” method has a “Fading mode” to combine data together in a blended way which can be helpful for sound or vibration data.
Another method called “Extract segment (blocks)” can be used to pull out statistics from frequency, order, and other non-time based functions.
Take for example the spectrum shown in Figure 29 below. Here the RMS (Root Mean Square) Value from 10 to 140 Hz is desired.
Figure 29: Double X cursor in Simcenter Testlab display shows the RMS value from 10 to 140 Hz.
While the RMS of a curve in a display can be quickly calculated using a Double X cursor, it might be desirable to calculate the RMS (from 10 to 140 Hz) on many data channels and export all the data to Excel.
This can be done in an automated way using the “Extract segment (block)” method with the process shown in Figure 30 below:
Figure 30: The “Extract Segments (Block)” method is part of the Interactive Analysis add-in. It can be used to calculate the RMS (or other statistics) over a limited frequency range.
To get the “Extract segment (block)” method, under “File -> Add-ins” turn on “Interactive Analysis” (16 tokens).
In the process, the “Spectrum average” method calculates a frequency spectrum from a time history. The frequency range from 10 Hz to 140 Hz is extracted, an RMS calculation is performed on the data from 10 Hz to 140 Hz, and the RMS data is sent to Excel.
To restrict the calculation to 10 Hz to 140 Hz, double click on the “Extract segment (block)” method and enter the “Segment start” and “Segment end” as shown in Figure 31:
Figure 31: Double click on the “Extract segment (block)” to set the range of data for the subsequent RMS calculation.
Note that the “Segment start” and “Segment end” require engineering units in addition to a number. In the case of a frequency spectrum, this is “Hz”. In the case of an order cut, this would be “rpm”.
6. Results
All methods output a new file containing only a segment of the original file. Example output results are shown in Figure 32 below:
Figure 32: The output of a segment extraction method is typically a folder called “Process” in a run (top picture). There may be links to the original throughput data as well (bottom picture).
The segment of data extracted is usually stored in a new run called “Process”. Often there is a link to the original data as well, for example a "Throughput" time file. When viewing data, highlighting the run may show both the original data and extracted data in the display. Highlight just the “Process” folder to see only the extracted data.
It is also possible to create a new run that contains only the extracted segment without a link to the original recording. This can be done by changing the “Input Data” setting from “Save time data as link” to “Do not save” in the “Input” method as shown in Figure 33 below:
Figure 33: To see only the extracted segment in the results, change the “Input Data” setting from “Save time data as link” to “Do not save” in the “Input” method properties.
An extracted segment of data can then be used for further processing by connecting it to additional methods (spectrums, overall level, etc).
