arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 1709.07089eess.SYcs.LGcs.SYstat.ML
收藏

关于为高效控制器学习设计LQR核

On the Design of LQR Kernels for Efficient Controller Learning

Max Planck Institute for Intelligent Systems(马克斯·普朗克智能系统研究所);Computational Learning and Motor Control Lab(计算学习与运动控制实验室)

Alonso Marco, Philipp Hennig, Stefan Schaal, Sebastian Trimpe

AI总结:

本文提出两种基于LQR结构的核,用于改进基于贝叶斯优化的控制器学习,通过模拟线性和非线性系统证明其优于传统GP方法。

摘要:

从数据中为非线性动态系统寻找最优反馈控制器是困难的。最近,贝叶斯优化(BO)被提出作为直接从实验试次中调整控制器的强大框架。为了选择下一个查询点并找到全局最优解,BO依赖于潜在目标函数的概率描述,通常为高斯过程(GP)。本文显示,使用常见核的GP在标准二次控制问题上可能导致学习效果差。对于一阶系统,我们构建了两种核,专门利用广为人知的线性二次调节器(LQR)的结构,同时保留贝叶斯非参数学习的灵活性。对不确定线性和非线性系统的模拟显示,LQR核在学习性能上优于传统GP方法。

英文摘要:

Finding optimal feedback controllers for nonlinear dynamic systems from data is hard. Recently, Bayesian optimization (BO) has been proposed as a powerful framework for direct controller tuning from experimental trials. For selecting the next query point and finding the global optimum, BO relies on a probabilistic description of the latent objective function, typically a Gaussian process (GP). As is shown herein, GPs with a common kernel choice can, however, lead to poor learning outcomes on standard quadratic control problems. For a first-order system, we construct two kernels that specifically leverage the structure of the well-known Linear Quadratic Regulator (LQR), yet retain the flexibility of Bayesian nonparametric learning. Simulations of uncertain linear and nonlinear systems demonstrate that the LQR kernels yield superior learning performance.

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。

↑