Enum UnitdentWarnings

Namespace

TimeSeriesAnalysis.Dynamic

Assembly

TimeSeriesAnalysis.dll

Enum of recognized warning or error states during identification of process model

public enum UnitdentWarnings

Fields

ClosedLoopEst_GlobalSearchFailedToFindLocalMinima = 15

If this is a closed loop system where setpoints of PID-controller change, and the "global search" at step1 of the ClosedLoopIdentifier failed to find a local minima when trying different gains. This warning likely means that the linear gain can be totally off the mark and may be too low.

ConstantInputU = 11

One or more inputs were constant and not identifiable, this can affect other parameters as well, and consider removing the input or using a different dataset

CorrelatedInputsU = 12

Correlated inputs

DataSetVeryShortComparedtoTMax = 1

The dataset is time span is very short compared to the maximal time-constant given as input to the algorithm

DynamicModelEstimationFailed = 16

UnitIdentify was able to solve the given estimation problem for the linear/static case, but when adding dynamics, the estimation failed for some reason.

NonCausalNegativeTimeConstant = 17

A negative time constant was returned, this is "non-causal", and could be due to a time-shift between u and y in the dataset.

NotPossibleToIdentify = 7

It was not possible to identify the

Nothing = 0

No errors or warnings

ReEstimateBiasDisabledDueToNonzeroDisturbance = 10

If disturbance is nonzero, then re-estimation of bias is turned off

ReEstimateBiasFailed = 9

Re-estimating bias method returned null, so the bias from the intial estimation is used, be careful the bias estimate may be off!

RegressionProblemFailedToYieldSolution = 2

RegressionProblemNaNSolution = 14

Some of the parameters returned were NaN, this can happen in the input and output vectors are all constants, such as all-zero

TimeConstantEstimateNotConsistent = 6

Time constant estimates vary significantly across dataset, indicating that something is wrong

TimeConstantEstimateTooBig = 8

Estimation returned an enourmous time constant, this is an indication of something is wrong

TimeConstantNotIdentifiable = 5

Time constant is not identifiable from dataset

TimeDelayAtMaximumConstraint = 3

The time delay which gave the lowest objective function is the biggest allowed time delay

• consider increasing this limit or if something is wrong

TimeDelayInternalInconsistencyBetweenObjFunAndUncertainty = 4

When considering different time delays internally, you expect the "best" to have both the lowest objective functino and the lowest paramter uncertainty

• but for some reason this is not the case