arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1503.05933math.OC
收藏

解耦系统的精确有效Hamilton-Jacobi可达性

Exact and Efficient Hamilton-Jacobi Reachability for Decoupled Systems

Mo Chen, Claire J. Tomlin

摘要:

可达性分析对于研究最优控制问题和微分对策具有重要意义,而微分对策是分析和建模机器人、飞行器控制等应用领域中许多实际问题的有力理论工具。在可达性分析中,人们对计算可达集感兴趣,可达集被定义为存在控制的状态集,尽管存在最坏的干扰,该控制可以将系统驱动到一组目标状态。根据应用程序,目标状态可用于对不安全或理想的配置进行建模。许多哈密尔顿-雅可比公式允许计算可达集;然而,由于计算时间和空间的指数复杂度缩放,涉及大约5维的问题变得棘手。在文献中存在许多计算近似解的方法,但是这些方法以复杂性换取最优性。在本文中,我们消除复杂性最优性权衡时不变的解耦系统使用解耦的Hamilton-Jacobi公式,使精确的重建高维解通过低维解的解耦子系统。我们的配方是兼容现有的数值工具,我们显示的准确性,计算的好处,我们的新方法使用两个数值例子的应用。

英文摘要:

Reachability analysis is important for studying optimal control problems and differential games, which are powerful theoretical tools for analyzing and modeling many practical problems in robotics, aircraft control, among other application areas. In reachability analysis, one is interested in computing the reachable set, defined as the set of states from which there exists a control, despite the worst disturbance, that can drive the system into a set of target states. The target states can be used to model either unsafe or desirable configurations, depending on the application. Many Hamilton-Jacobi formulations allow the computation of reachable sets; however, due to the exponential complexity scaling in computation time and space, problems involving approximately 5 dimensions become intractable. A number of methods that compute an approximate solution exist in the literature, but these methods trade off complexity for optimality. In this paper, we eliminate complexity-optimality trade-offs for time-invariant decoupled systems using a decoupled Hamilton-Jacobi formulation that enables the exact reconstruction of high dimensional solutions via low dimensional solutions of the decoupled subsystems. Our formulation is compatible with existing numerical tools, and we show the accuracy, computation benefits, and an application of our novel approach using two numerical examples.

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。

↑