{"id":"arxiv/2603.24489","url":"https://arxiv.org/abs/2603.24489","tags":["math.OC","eess.SY"],"score":0,"title":"Model Predictive Path Integral Control as Preconditioned Gradient Descent","domain":"math.OC","source":"arxiv","authors":["Mahyar Fazlyab","Sina Sharifi","Jiarui Wang"],"reviews":0,"subject":"math.OC","summary":"Model Predictive Path Integral (MPPI) control is a widely used sampling-based method for trajectory optimization, yet its convergence properties remain only partially understood. This paper provides a direct convergence analysis using variational optimization. By lifting constrained trajectory optimization to a Kullback-Leibler (KL) regularized problem over decision distributions, we derive a reduced free-energy objective defined over a parametric sampling family. For general parametric families, we derive gradient and Hessian representations of this reduced objective and analyze…","category":"research","comments":0,"disputed":false,"division":"math","mentions":0,"discussed":0,"published":"2026-03-25","reactions":{}}