{"id":"arxiv/2607.08942","url":"https://arxiv.org/abs/2607.08942","tags":["eess.SY","cs.RO","math.OC"],"score":0,"title":"Adaptive MPPI with Online Disturbance Covariance Estimation: Provable Stability Tightening via Spatial Smoothing","domain":"eess.SY","source":"arxiv","authors":["Hyung-Jin Yoon","Hunmin Kim"],"reviews":0,"subject":"eess.SY","summary":"We study Model Predictive Path Integral (MPPI) control for nonlinear systems with additive process disturbances whose covariance is unknown, spatially varying, and slowly time-varying. A mismatched disturbance covariance produces a persistent penalty in closed-loop stability certificates, while online estimation can reduce this penalty as data are collected. We propose a cell-wise recursive covariance estimator with spatial diffusion and prove a finite-horizon error bound that separates stochastic-approximation error, spatial-smoothing bias, and temporal-drift effects. The diffusion kernel…","category":"research","comments":0,"disputed":false,"division":"eess","mentions":0,"discussed":0,"published":"2026-07-09","reactions":{}}