Task Decomposition for Iterative Learning Model Predictive Control

Task Decomposition for Iterative Learning Model Predictive Control
复制标题

迭代学习模型预测控制的任务分解

DOI:
10.23919/acc45564.2020.9147625
复制
发表时间:
2019
期刊:
2020 American Control Conference (ACC)
影响因子:
--
通讯作者:
F. Borrelli
F. Borrelli
中科院分区:
--
文献类型:
--
作者:
Charlott Vallon;F. Borrelli

文献摘要

参考文献

被引文献

相似文献

提出了一种迭代学习模型预测控制的任务分解方法。我们考虑一个受约束的非线性动力系统,并假设状态输入对数据集的可用性,解决了任务$\mathcal{T}1$。我们的目标是找到一个可行的模型预测控制策略的第二个任务,$\mathcal{T}2$,使用存储的数据从$\mathcal{T}1$。我们的方法适用于任务$\mathcal{T}2$,这些任务由$\mathcal{T}1$中包含的子任务组成。在本文中,我们提出了子任务和任务分解问题的形式化定义,并提供了可行性证明和相对于简单初始化的迭代成本改进。我们证明了所提出的方法的有效性自主赛车和机器人操作实验。
A task decomposition method for iterative learning model predictive control is presented. We consider a constrained nonlinear dynamical system and assume the availability of state-input pair datasets which solve a task $\mathcal{T}1$. Our objective is to find a feasible model predictive control policy for a second task, $\mathcal{T}2$, using stored data from $\mathcal{T}1$. Our approach applies to tasks $\mathcal{T}2$ which are composed of subtasks contained in $\mathcal{T}1$. In this paper we propose a formal definition of subtasks and the task decomposition problem, and provide proofs of feasibility and iteration cost improvement over simple initializations. We demonstrate the effectiveness of the proposed method on autonomous racing and robotic manipulation experiments.
DOI: 10.1109/tcst.2019.2912345
发表时间: 2020-07-01
影响因子: 4.8
作者:
Cobb, Mitchell K.;Barton, Kira;Vermillion, Chris
通讯作者: Vermillion, Chris