Model reduction

This is an overview of the functions available for model reduction.

State reduction

Function	Description
lpvmodred ('ltibalred' method)	Balanced truncation of LTI part of LPVcore.lpvlfr model
lpvmodred ('momentmatching' method)	Moment matching of discrete-time LPVcore.lpvss models given a maximum state order
lpvmmred	Moment matching of discrete-time LPVcore.lpvss models given a matching window

Scheduling reduction

Note that the scheduling reduction approaches always return an LPVcore.lpvss object with affine dependence on the reduced scheduling signal - even if the original dependence if non-affine.

Function	Description
lpvpcared	Scheduling reduction based on Principal-Component Analysis (PCA)
lpvddnred	Scheduling reduction based on Deep Neural Network (DNN). With support for LPVcore.lpvgridss objects.

$\Delta$-block reduction

Function	Description
lpvmodred ('deltasvdred' method)	SVD-based reduction of $\Delta$-block
lpvmodred ('deltabalred' method)	Balanced reduction of $\Delta$-block