arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2105.07693cs.LGcs.ROcs.SYeess.SY
收藏

通过近似贝叶斯输入推断实现高效随机最优控制

Efficient Stochastic Optimal Control through Approximate Bayesian Input Inference

Technical University of Darmstadt(达姆施塔特工业大学);Aalto University(阿尔托大学);Finnish Centre of AI(芬兰人工智能中心)

Joe Watson, Hany Abdulsamad, Rolf Findeisen, Jan Peters

AI总结:

本文将随机最优控制重构为贝叶斯输入估计问题,提出基于近似推断的高斯求解器,在非线性仿真任务中优于常用基线。

摘要:

不确定性下的最优控制因诸多原因仍是一项普遍挑战。关键困难之一在于为底层随机优化问题生成易处理的解。我们展示了如何通过将控制问题构建为输入估计问题,利用先进的近似推断技术,以有原则且实用的方式处理统计近似。通过分析高斯情形,我们提出了一种基于推断的求解器,它在随机和确定性环境中均有效,并且在非线性仿真任务上被发现优于常用基线。我们建立了这种推断表述与以往随机最优控制方法之间的联系,并概述了该推断视角因其统计本质而带来的若干优势。

英文摘要:

Optimal control under uncertainty is a prevailing challenge for many reasons. One of the critical difficulties lies in producing tractable solutions for the underlying stochastic optimization problem. We show how advanced approximate inference techniques can be used to handle the statistical approximations principled and practically by framing the control problem as a problem of input estimation. Analyzing the Gaussian setting, we present an inference-based solver that is effective in stochastic and deterministic settings and was found to be superior to popular baselines on nonlinear simulated tasks. We draw connections that relate this inference formulation to previous approaches for stochastic optimal control and outline several advantages that this inference view brings due to its statistical nature.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑