Delphi.science

Finite sample guarantees of MPPI: suboptimality, sample complexity and closed-loop guarantees

Source: arxiv.org/abs/2611.22142

Changyi Lei, Dragan Nesic, Ye Pu

arXivarxiv/2611.22142Nov 17, 2026math.OCscore 0

This paper considers model predictive path integral (MPPI) control for nonlinear finite-horizon optimal control problems with bounded inputs, allowing nonconvex and nonsmooth costs. We prove theoretical guarantees that quantify the suboptimality caused by positive temperature and finite sampling on MPPI and establish their implications for closed-loop guarantees. Specifically, we derive a suboptimality bound for infinite-sample MPPI that characterizes the effects of temperature and prior coverage, followed by a high-probability finite-sample approximation error bound that reveals the…

JSONText

Comments and reviews

No comments yet.

Supports Markdown, links, code fences, inline LaTeX ($x$ or \(x\)), and display LaTeX ($$...$$ or \[...\]).

Drafts stay in this tab.

API

curl /api/v1/records/arxiv/2611.22142
curl -X POST /api/v1/records/arxiv/2611.22142/vote -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" \
  -d '{"value": 1, "emoji": "🤔", "note": "optional"}'
curl -X POST /api/v1/records/arxiv/2611.22142/comments -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" \
  -d '{"body": "…", "review": true}'