IEEE PES Schenectady Chapter Virtual Lecture “Data-Driven Modeling of IBR: Black-Box and White-Box Approaches”

#data-centers #data-driven-modeling #power-systems #IBR

Title: Data-Driven Modeling of Inverter-Based Resources: Black-Box and White-Box Approaches

Abstract:

As inverter-based resources (IBRs) play a growing role, accurate dynamic models are increasingly critical for reliable planning and operation of power systems highly penetrated by renewables, storage, and data centers. In practice, however, IBR modeling is challenging because some resources come with imperfect or outdated engineering models, while others appear as black boxes due to proprietary controls and undocumented mode changes. This webinar introduces two complementary data-driven approaches to this problem: white-box methods that use measurements to validate and calibrate known models, and black-box methods that learn dynamic behavior directly from noisy data when internal model details are unavailable. The webinar emphasizes the practical value of combining both perspectives. White-box approaches offer interpretability and support model validation when structure is known, whereas black-box approaches enable faithful modeling when the resource is inaccessible from a modeling standpoint. Together, they form a flexible foundation for improving IBR representation and strengthening grid stability studies in diverse practical settings.