CygNet Measurement > FMS Explorer > Admin Menu > Managing Estimation Engines

Managing Estimation Engines

Note: This functionality is only available in systems licensed for FULL mode.

Estimation Use the Estimation button on the Admin menu to access options to manage estimation engines and rules. The CygNet Measurement system supports the configuration of multiple estimation engines and rules. Estimation engines define when and what types of data will be estimated, and estimation rules define how to perform those estimations to arrive at better data values. You can add, view or edit, or delete estimation engines and rules via this interface. Changes made to estimation engines and rules are historized in the measurement system.

Estimation is only performed at the station level; it is not meaningful to estimate a meter run.

Data will be estimated when any of the following is true.

Each estimation engine can be configured to contain multiple estimation rules, so that different estimation handling routines can be performed in a user-defined order of precedence, given data constraints. 

Missing data and low quality data is estimated as soon as the record enters the measurement system. Late data estimation must be initiated by executing the Estimate Station Data command.

You must have proper security authorization to perform these actions. See FMS Security (and ACCESS security event) for information about configuring security access to estimation engines and rules.

Managing Estimation Engines

Estimation Engines are configured via the Admin menu in FMS Explorer. Click Estimation on the Admin ribbon to access the Estimation Engines dialog box.

Estimation Engines

The Estimation Engines dialog box lists all defined estimation engines in the measurement system. Each row displays a configured estimation engine name, the type(s) of data to which it applies, the active begin and end times, and whether or not the engine is enabled. Results can be filtered by the values of any column. You can also access options to add, view or edit, or delete estimation engines and rules from this dialog box.

Creating Estimation Engines

Click Add to create a new estimation engine in your system. The Create New Estimation Engine dialog box will appear.

Create New Estimation Engine

The properties for the Create New Estimation Engine dialog box are as described for the Estimation Engine Properties dialog box below, with the exception of the historization navigator.

Back to top

Viewing or Editing Estimation Engines

Click an estimation engine name and then click Edit to view or modify the properties for an existing estimation engine, or to view its historized records. The Estimation Engine Properties dialog box will appear.

Estimation Engine Properties

Estimation Engine Properties

The properties for the Estimation Engine Properties dialog box are described below. This dialog box is similar to the Create New Estimation Engine dialog box above.

Parameter Description

Name

The name of the estimation engine.

Enabled

Click to select/clear whether or not to enable this estimation engine.

Node

The name of the Node to which this estimation engine applies. Click ... to select a Node.

Note: When choosing a group Node, it must be a General Group Node of uniqueness type: Estimation Engine.

Estimation categories

Estimate missing data

Select to estimate missing data when there are gaps in data that are not likely to be filled from the device.

Estimate late data

Select to estimate late data. Data is late if it has not entered the system within a user-defined interval.

Hours before late

Type the number of hours after which late data is considered "late" and will be estimated.

Estimate low quality data

Select to estimate or override data that is at or below a user-defined data quality threshold.

Low data quality

Select a low data quality threshold from the drop-down menu. Records with data quality at or below the selected setting will be estimated.

System data quality settings are configured on the Administrative Options dialog box. See Configuring Data Quality Settings for more information.

Use station's low quality

Select to use the low end of the station's data quality range.

For stations with multiple devices of varying data quality, the lowest device data quality value will be used to determine estimation for the station.

Use station's high quality

Select to use the high end of the station's data quality range.

For stations with multiple devices of varying data quality, the highest device data quality value will be used to determine estimation for the station.

Estimation Rules

Estimation Rules

A list of all the estimation rules that have been defined for this engine.

Columns include the following.

  • #
  • Rule
  • Assigned Data Quality
  • Minimum Data Quality
  • Hours Back

#

Estimation rules will be performed in the specified order of precedence. Click Move up and Move down to reorder the rules.

Rule

The type of estimation rule. See Add Rule below for a list of estimation rule types and their descriptions.

Assigned Data Quality

The data quality level that will be assigned to the resultant estimated record, once it is estimated by the selected estimation rule. Click the down arrow to select a Data Quality value from the drop-down menu.

Min Data Quality

The minimum data quality level for other records that are used as part of the estimation, for the selected estimation rule.

Example

If the data quality for the last known value is higher than the assigned data quality for the estimated record, use the higher data quality value.

Click the down arrow to select a Data Quality value from the drop-down menu.

Hours Back

For historical rules, the number of hours into the past to use to estimate a historical average.

Add Rule

Click Add Rule to add a new estimation rule to the estimation engine. Select a rule type from the drop-down menu.

Note: For estimating non-contiguous data, the estimated value can only be determined from data that already exists in the database (records that have previously been saved).

Options are as follows.

  • Last good value — The estimation process will use the most recent "good value" (value with a data quality above the minimum data quality configured for the rule) as the estimated value.
  • Historical — The estimation process will estimate data using a historical average of previous data, for a time period defined in the rule.
  • Interpolation — The estimation process will use an average of good data from before and after the data to be estimated. Interpolation can be used to estimate missing data, and to estimate data of low quality (assuming there is a good record before and after).

    Example

    If volume data is missing for three hours and the last good record is 100 MCF, and the next good record is 200 MCF, then the first estimated record would be 125 MCF, the second 150 MCF, and the third 175 MCF.

  • Custody/Check — The estimation process will copy the data for a configured “linked station” for the time period being estimated.
  • Same hour — The estimation process will find values from the same hour of a previous day to fill in missing data. To be used, the data must be from the same hour within the day as the missing data, and must meet a specified data quality. If data for the same hour of the previous day is not of sufficient quality, the prior day will be checked, and so on, until an acceptable value is found.

    Note: If the maximum time is exceeded without finding an acceptable same hour value, a warning will be generated in the FMS command log and the estimation process will proceed without estimating this value.

Delete

Click Delete to delete a selected estimation rule.

Move up / Move down

Select any rule and then click Move up or Move down to change the rule execution order.

Details

Select any rule and then click Details to access the Record Details for the selected estimation rule.

Historization Navigator

Record details

Where applicable, click Record details in the historization navigator to view sequential historized records for the selected estimation engine.

Audit history

Where applicable, click Audit history in the historization navigator to view detailed history of audit trail data changes for the selected estimation engine.

Back to top

Deleting Estimation Engines

Use the Delete button to either delete an Estimation Engine from a point in time, or to completely purge an Estimation Engine from the system. You can only take this action for one estimation engine at a time.

Click an estimation engine name and then click Delete to delete or purge the selected estimation engine. The Delete Estimation Engine dialog box will appear.

Delete Estimation Engine

You will need to select whether to delete the estimation engine from a point in time, or to completely purge the estimation engine from the system. If you select to Delete from a Point in Time, also specify the PIT (C) using the date/time selector provided. See Using the Home Menu for more information about PIT.

Note: Be certain of your selection, as there is NO "undo" for this action.

A message will appear asking you to confirm that you wish to delete/purge the selected Estimation Engine from your system. Select Yes or No.

Back to top

Historization Navigator

The Estimation Engine Properties dialog box contains a historization navigator. Use the historization navigator to access sequential historized records for the same estimation engine, where applicable.

Historization Navigator

The following fields are displayed in the historization navigator.

Field Description

<<

Click << to navigate to the previous historized record, if applicable.

[Begin date]

Displays the effective begin date, in Contract time (C), of the historized record currently selected.

[End date]

Displays the effective end date, in Comntract time (C), of the historized record currently selected.

>>

Click >> to navigate to the next historized record, if applicable.

Record details

Use the << and >> navigation buttons to select a historized record date span, in Contract time (C), for which you want to view record details. Next, click Record details to view details of the record such as who made the change, the date and time it was made, and optionally the audit category and/or any comments entered regarding the change.

See Viewing Record Details for more information.

Audit history

Click Audit history to view detailed history of audit trail value changes for the Node in its entirety.

See Viewing Audit History for more information.

Back to top

Saving FMS Record Changes

After creating or editing an Estimation Rule, you'll be asked to save your changes via the Save FMS Record dialog box. Configure the date and audit information requested to historize the changes in your system.

Save FMS Record

Property Description

Change begin

Set the effective date and time for the record change to begin.

Options are as follows.

  • Indefinite: Click to make the record change begin date and time effective for an undefined amount of time. Effectively, it means all time.
  • Choose date/time: Click to make the record change begin date and time effective for a specific date and time. Use the date/time selector to select the specific begin date and time.

Change end

Set the effective date and time for the record change to end.

Options are as follows.

  • Indefinite: Click to make the record change end date and time effective for an undefined amount of time. Effectively, it means all time.
  • Choose date/time: Click to make the record change end date and time effective for a specific date and time. Use the date/time selector to select the specific end date and time.

Audit category

Select an appropriate Audit category from the drop-down menu. The audit category type for this kind of edit is Configuration Change.

Audit categories are configured on the Admin menu. See Configuring Audit Categories for more information.

Comments

Type Comments regarding creation or modification of the rule or Node in the text box. Although comments are optional, they become part of the record and can be helpful when auditing or reviewing records.

Back to top

Let us know how we can improve this topic.

CygNet at weatherford.com

© 2020 Weatherford. All rights reserved.