Skip to main content

AuxPredMultiStep class

Stores predictor model properties required specifically for multistep identification.

Construction​

  • aux = AuxPredMultiStep()  creates an empty AuxPredMultiStep object

Properties​

  • P​

    PredictorMap object

    Predictor mapping from reconstructed innovations ( ξ^\hat{\xi} ) to output node.

  • arx​

    structure array

    Input/output sets for first step (estimating high-order ARX model to reconstruct innovations) of the multistep algorithm. Must contain the following arrays of indices referring to nodes/excitations in the underlying network structure of the predictor model:

    • Uset: Input nodes
    • Yset: Output nodes
    • Rset: Input excitations

  • xi​

    Integer array

    Reconstructed innovation signals included in the final predictor model, specified by indices referring to nodes in the underlying network structure. This property must be a subset of arx.Yset.

Methods​