AcceptFilter

General Mission Analysis Tool

previous page next page

AcceptFilter

AcceptFilter — Allows selection of data subsets for processing by the batch least squares estimator.

Description

Starting with the R2017A release of GMAT, the AcceptFilter resource replaces the StatisticsAcceptFilter resource. The StatisticsAcceptFilter resource is still available in this release but it is deprecated and will be removed in a future release.

The AcceptFilter object is used to create criteria for the inclusion of subsets of the available data in the estimation process based on observation frequency, tracker, measurement type, record number, or time. Instances of AcceptFilter are specified for use on the DataFilters field of a TrackingFileSet or BatchEstimatorInv object.

GMAT implements two levels of data editing for estimation. First-level editing criteria are specified on the DataFilters field of the TrackingFileSet instance. At this level, the user may choose what data is admitted into the overall pool of observations provided to the estimator. Any data excluded at the tracking file set level will be immediately discarded and not available to the estimation process.

Second-level data editing is specified on the DataFilters field of the BatchEstimatorInv instance. At this level, the user may choose what data is used in the estimation state update. Residuals will be computed for any observations admitted through first-level editing, but any data excluded at the estimator level will be flagged as user edited, and will not affect the computation of the state correction. This allows the user to evaluate the quality of untrusted data against a solution computed using a trusted set of measurements.

A single AcceptFilter may employ multiple selection criteria (for example simultaneously thinning different stations or data types by differing intervals). Multiple criteria on a single filter are considered in an AND sense. When multiple criteria are specified on a single filter, an observation must meet all specified criteria to be accepted.

Multiple AcceptFilters with different selection criteria may be specified on a single TrackingFileSet or BatchEstimatorInv. When multiple filters are specified, these act in an OR sense. Data meeting criteria for any of the specified filters will be accepted.

Fields

Field

Description

DataTypes

List of data types

Data Type	String Array
Allowed Values	A set of any supported GMAT measurement types, or 'All'
Access	set
Default Value	{All}
Units	N/A
Interfaces	script

EpochFormat

Allows user to select format of the epoch

Data Type	String
Allowed Values	UTCGregorian, UTCModJulian, TAIGregorian, TAIModJulian, TTGregorian, TTModJulian A1Gregorian, A1ModJulian, TDBGregorian, TDBModJulian
Access	set
Default Value	TAIModJulian
Units	N/A
Interfaces	script

FileNames

List of file names (a subset of the relevant TrackingFileSet's FileName field) containing the tracking data. If this field equals From_AddTrackingConfig, then two things happen; (1) All of the files in the relevant TrackingFileSet are used as a starting point, and (2) Of the data in all of the files, only the data defined by the AddTrackingConfig field of the relevant TrackingFileSet are used. This field is only applicable when the AcceptFilter is used on a TrackingFileSet.

Data Type	StringArray
Allowed Values	valid file name, 'All', or 'From_AddTrackingConfig'
Access	set
Default Value	{All}
Units	N/A
Interfaces	script

FinalEpoch

Final epoch of desired data to process

Data Type	String
Allowed Values	any valid epoch
Access	set
Default Value	latest day defined in GMAT
Units	N/A
Interfaces	script

InitialEpoch

Initial epoch of desired data to process

Data Type	String
Allowed Values	any valid epoch
Access	set
Default Value	earliest day defined in GMAT
Units	N/A
Interfaces	script

ObservedObjects

List of user-created tracked objects (e.g., name of the Spacecraft resource being tracked)

Data Type	Object Array
Allowed Values	User defined observed object or 'All'
Access	set
Default Value	{All}
Units	N/A
Interfaces	script

RecordNumbers

A list of one or more single record numbers or spans of record numbers to accept. Observation record numbers are reported in the GMAT estimator output file. This field is only applicable when the AcceptFilter is used on the estimator level.

Data Type	String array
Allowed Values	Integers or spans of integers (see examples)
Access	set
Default Value	{All}
Units	N/A
Interfaces	script

ThinMode

'Frequency' for record count frequency mode and 'Time' for time interval mode. This field is only applicable when the AcceptFilter is used on a TrackingFileSet.

Data Type	String
Allowed Values	'Frequency' or 'Time'
Access	set
Default Value	Frequency
Units	N/A
Interfaces	script

ThinningFrequency

If ThinMode is Frequency, the integer 'n' is used to specify that every nth data point should be accepted. For example, 3 specifies that every third data point, meeting all the accept criteria, should be accepted and 1 specifies that every data point, meeting all the accept criteria, should be accepted. If ThinMode is Time, the integer 'n' is a number of seconds between accepted observations, using the first available observation as the anchor epoch. For example, a value of 300 means that observations will be accepted every 300 seconds, starting from the first available observation. This field is only applicable when the AcceptFilter is used on a TrackingFileSet.

Data Type	Integer
Allowed Values	Positive Integer
Access	set
Default Value	1
Units	Depends on ThinMode value
Interfaces	script

Trackers

List of user-created trackers (e.g., name of the GroundStation resource being used)

Data Type	Object Array
Allowed Values	any valid user-created Tracker object (e.g., GroundStation) or 'All'
Access	set
Default Value	{All}
Units	N/A
Interfaces	script

Remarks

Some fields of AcceptFilter are not applicable at either the first-level (tracking file set) or second-level (estimator) editing stages. The RecordNumbers field has no functionality when applied to an accept filter at the tracking file set level. The FileNames, ThinningFrequency, and ThinMode fields have no functionality when applied to an accept filter at the estimator level.

Use of combinations of instances of AcceptFilter and RejectFilter at both levels is permitted.

Examples

First-level (TrackingFileSet) Data Editing

The following examples illustrate use of an AcceptFilter for first-level data editing. At this level, the AcceptFilter instance should be assigned to the DataFilters field of a TrackingFileSet. In these examples, only data meeting the criteria specified by the accept filter will be admitted through. All other data is immediately discarded.

This example shows how to create an AcceptFilter to sample the data at a frequency of 1:10 (thinning the data to one tenth of its volume).

Create AcceptFilter af;
  
af.ThinningFrequency = 10;

Create TrackingFileSet estData;

estData.DataFilters = {af};

BeginMissionSequence;

The next example will accept all data from station GDS and accept every 5th observation from station CAN. Only data from stations GDS and CAN will be accepted.

Create AcceptFilter af1;
Create AcceptFilter af2;
 
Create GroundStation GDS CAN;

af1.Trackers          = {'GDS'}; 
af2.Trackers          = {'CAN'};
af2.ThinningFrequency = 5;
 
Create TrackingFileSet estData;
 
estData.DataFilters = {af1, af2};

BeginMissionSequence;

The last example illustrates thinning data by time interval, using a 300-second thinning interval.

Create AcceptFilter saf;
 
af.ThinMode          = 'Time'; 
af.ThinningFrequency = 300;
 
Create TrackingFileSet estData;
 
estData.DataFilters = {af};

BeginMissionSequence;

Second-level (estimator) Data Editing

The following examples illustrate use of an AcceptFilter for second-level data editing. At this level, the AcceptFilter instance should be assigned to the DataFilters field of a BatchEstimatorInv. In these examples, only data meeting the criteria specified by the accept filter will be used in the estimation state update. Residuals will be computed for all available data (all data admitted at the first level), but data not accepted at the estimator level will be flagged as user edited.

This example shows how to create an AcceptFilter to accept specific data records by record number.

Create AcceptFilter af;
  
af.RecordNumbers = {10, 11, 20-150, 155-300};

Create BatchEstimatorInv bls;

bls.DataFilters = {af};

BeginMissionSequence;

The next example will accept only range data from station MAD over the time span 10 Jun 2012 02:56 to 13:59.

Create AcceptFilter af;
Create GroundStation MAD;

af.Trackers     = {'MAD'};
af.DataTypes    = {'Range'};
af.EpochFormat  = UTCGregorian;
af.InitialEpoch = '10 Jun 2012 02:56:00.000';
af.FinalEpoch   = '10 Jun 2012 13:59:00.000';

Create BatchEstimatorInv bls;

bls.DataFilters = {af};

BeginMissionSequence;

The last example illustrates accepting all data from station MAD and only range data from station CAN.

Create AcceptFilter af1 af2;
Create GroundStation MAD CAN;
 
af1.Trackers         = {'MAD'}; 
af2.Trackers         = {'CAN'};
af2.DataTypes        = {'Range'};
 
Create BatchEstimatorInv bls;
 
bls.DataFilters = {af1, af2};

BeginMissionSequence;

previous page start next page