arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2602.23313eess.SYcs.SY
收藏

基于深度可达性分析和分层控制架构的信号时间逻辑验证与综合

Signal Temporal Logic Evaluation and Synthesis Using Deep Reachability Analysis and Layered Control Architecture

School of Aeronautics and Astronautics, Purdue University(普渡大学航空与航天学院);Department of Mechanical Engineering, Ulsan National Institute of Science and Technology(蔚山科学技术院机械工程系);Interdisciplinary Engineering, Purdue University(普渡大学跨学科工程)

Joonwon Choi, Kartik Anand Pant, Youngim Nam, Henry Hellmann, Karthik Nune, Inseok Hwang

AI总结:

本文提出了一种基于深度可达性分析和分层控制架构的信号时间逻辑验证与综合框架,通过高效计算后向可达管和结合混合整数线性规划与模型预测控制,实现安全可靠的任务执行。

摘要:

我们提出了一种基于信号时间逻辑(STL)的框架,该框架严格验证了由STL描述的任务可行性,并综合控制以安全执行。该框架通过两个阶段确保安全可靠的运行。首先,该框架通过计算后向可达管(BRT)来评估STL的可行性,该BRT捕捉了所有能满足给定STL的状态,无论初始状态如何。该框架还针对更一般的STL规范解决了多可达避(MRA)问题,并利用深度神经网络来减轻可达性分析的计算负担,将计算时间减少了约1000倍相比基线方法。我们进一步提出了一种分层规划和控制架构,该架构结合混合整数线性规划(MILP)进行全局规划,并利用模型预测控制(MPC)作为验证STL的本地控制器。因此,该框架能够稳健地处理未在环境信息或STL中描述的障碍物的意外行为,从而提供可靠的任务性能。我们的数值模拟表明,该框架能够成功计算给定STL的BRT并执行任务。

英文摘要:

We propose a signal temporal logic (STL)-based framework that checks the feasibility of a mission described in STL and synthesizes control to safely execute it. The proposed framework evaluates the safety and reliability of operation through two phases. First, the proposed framework assesses the feasibility of STL by computing a backward reachable tube (BRT), which captures all states that can satisfy the given STL. The proposed framework accommodates the multiple reach-avoid (MRA) problem to address more general STL specifications and leverages a deep neural network to alleviate the computation burden for reachability analysis, significantly reducing the inference time compared to a baseline method. We further propose a layered planning and control architecture that combines mixed-integer linear programming (MILP) for global planning with model predictive control (MPC) as a local controller for the given STL. Consequently, the proposed framework can robustly handle unexpected behavior of obstacles that are not described in the environment information or STL, thereby providing reliable mission performance. Our numerical simulations demonstrate that the proposed framework can successfully compute BRT for a given STL and perform the mission.

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。

↑