LPV Modeling of Clusters in Dynamic Power System Models

This paper presents a systematic approach for turning the external area of a nonlinear power system model into a set of linear parameter-varying (LPV) models. The aim is to extend the scope of linear input-output equivalencing techniques by including some tunable parameters. In particular, the LPV models are designed to cover the changes in operating conditions that occur following the shutdown of generators. After partitioning the external area into multiple clusters, the dynamic behavior of a cluster is aligned to a new situation by scaling the inertia constant of its generators. The construction of a LPV model requires the assessment of the related linear cluster model for different values of the scale factor. A polynomial structure is subsequently fitted in the least-squares sense with respect to the collection of linear models. The quality of the LPV modeling approach is assessed on the IEEE 118-bus system model by studying multiple ways to disconnect generators from a cluster.