Wiki

Options

New Case Case Status
Log In

Wiki

Options

 
Knowledge Base Articles»KB100015 - Data Smoothing
  • RSS Feed

Last modified on 12/5/2018 5:04 PM by User.

Tags:

KB100015 - Data Smoothing

Introduction

Feature will be available in 10.4

Data smoothing is based on Exponentially Weighted Moving Average (EWMA) which is smoothing a time series set of data using the exponential window function.

Important Notes:

  1. The expected source value of an acquired point must be linear; Angular values (i.e. degrees) are not supported.
  2. If a source value of an acquired point has a bad quality (per OPC Quality standard), then all source values within the window will be marked as uncertain quality.
  3. While the quality is uncertain, the source values will not be smoothed unless the interface is configured to allow uncertain quality to be considered good.

 

Setup

To setup an acquired point with data smoothing, Open the Point Configuration form, then on the Function Setup tab, set the DampenWindow to a whole value between 0 and 86,400 seconds; If the value is set to 0, then smoothing is not active.

NOTE: The smoothing window in the Exponentially Weighted Moving Average is determined by DampenWindow divided by number of seconds from the Interfaces Refresh Rate (i.e.: DampenWindow of 60 at a 5 second refresh rate would have a window size of 12 samples), a DampenWindow that results in a remainder, the remainder will be removed (i.e.: DampenWindow of 60 at an 8 second refresh rate would have a window size of 7 samples).

To reduce the number of acquired points stored in the archive, ExceptionMinimumTime must be greater than 0. ExceptionMinimumTime is the minimum time span between exceptions, expressed in seconds. To store the same amount of acquired points prior to the new refresh rate, set the ExceptionMinimumTime to old refresh rate (in seconds).

 

 

If a quality is uncertain and the values are to be included in the averaging in the Exponentially Weighted Moving Average (EWMA),

Set Include Uncertain to True

 

Example

In this example, the acquired point is to be smoothed over a 5 minute window of time with a refreshed source value updated every 15 seconds.

DampenWindow = 5 minutes * 60 seconds = 300 seconds

Refresh Rate = 15 seconds

Result is a window size of 20 samples

 

In EtaPRO Data Interface Configurator, setup the interface as below:

Include Uncertain: False

Refresh Rate:  15 seconds

 

In the Point Configuration form, then on the Function Setup tab, Set the following:

DampenWindow: 300

Refresh Rate ID: 1

 

Now, every source value acquired will now be an exponentially smoothed value.

 

Converting an Existing Point to Include Data Smoothing Example

In this example, the acquired point is to be smoothed over a 1 minute window of time with a refreshed source value updated every 5 seconds. The source value is to be saved to the database at minimum of 1 minute.

DampenWindow = 1 minute * 60 seconds = 60 seconds

Refresh Rate = 5 seconds

ExceptionMinimumTime = 60 seconds

Result is a window size of 12 samples

 

STEP: In EtaPRO Data Interface Configurator, Set the following:

Refresh Rate: 5 seconds

 

STEP: In EtaPRO Point Configuration form, then on the Function Setup tab, Set the following:

DampenWindow: 60

ExceptionMinimumTime: 60