processMonitor {mvMonitoring}

Adaptive Process Training

Description

Apply Adaptive-Dynamic PCA to state-specific data matrices.

Usage

processMonitor(
  data,
  trainObs,
  updateFreq = ceiling(0.5 * trainObs),
  faultsToTriggerAlarm = 5,
  ...
)

Arguments

data

An xts data matrix

trainObs

The number of training observations to be used

updateFreq

The number of non-flagged observations to collect before the function updates. Defaults to half as many observations as the number of training observations.

faultsToTriggerAlarm

The number of sequential faults needed to trigger an alarm. Defaults to 5.

...

Lazy dots for additional internal arguments

Details

This function is the class-specific implementation of the Adaptive- Dynamic PCA described in the details of the mspTrain() function. See the mspTrain() function's help file for further details.

This internal function is called by mspTrain(). This function calls the faultFilter() function.

Value

A list with the following components:

FaultChecks –

a class-specific xts flagging matrix with the same number of rows as "data". This flag matrix has the following five columns:

SPE –

the SPE statistic value for each observation in "data"

SPE_Flag –

a vector of SPE indicators recording 0 if the test statistic is less than or equal to the critical value passed through from the threshold object

T2 –

the T2 statistic value for each observation in "data"

T2_Flag –

a vector of T2 fault indicators, defined like SPE_Flag

Alarm –

a column indicating if there have been five flags in a row for either the SPE or T2 monitoring statistics or both. Alarm states are as follows: 0 = no alarm, 1 = Hotelling's T2 alarm, 2 = Squared Prediction Error alarm, and 3 = both alarms.

Non_Alarmed_Obs –

a class-specific xts data matrix of all the non-alarmed observations (observations with alarm state equal to 0)

Alarms –

a class-specific xts data matrix of the features and specific alarms of Alarmed observations, where the alarm codes are listed above

trainSpecs –

a threshold object returned by the internal threshold() function. See the threshold() function's help file for more details.

See Also

Calls: faultFilter. Called by: mspTrain.

Examples

nrml <- mspProcessData(faults = "NOC")
data <- nrml[nrml[,1] == 1]

processMonitor(data = data[,-1], trainObs = 672)