{"id":"arxiv/2607.10893","url":"https://arxiv.org/abs/2607.10893","tags":["eess.SY","math.OC"],"score":0,"title":"Streaming Contraction Certificates for Nonlinear Networks: Topology-Aware Data Sufficiency with Partial Observation","domain":"eess.SY","source":"arxiv","authors":["Faegheh K. Moazeni"],"reviews":0,"subject":"eess.SY","summary":"Certifying the safety of a control action in real time, from streaming partial observations of a nonlinear, interconnected system under non-stationary disturbances, is a problem no existing data-driven framework solves. Batch methods such as data-enabled predictive control require a pre-collected dataset and offer no stability certificate for nonlinear dynamics; informativity-based approaches characterize data sufficiency offline and non-recursively; neither exploits the known graph topology of networked systems as a structural prior. This paper addresses both limitations. First, we develop…","category":"research","comments":0,"disputed":false,"division":"eess","mentions":0,"discussed":0,"published":"2026-07-12","reactions":{}}