计算机科学
功能(生物学)
间断(语言学)
障碍物
放松(心理学)
控制(管理)
控制理论(社会学)
集合(抽象数据类型)
生命关键系统
控制工程
工程类
数学
软件
人工智能
数学分析
法学
程序设计语言
生物
进化生物学
社会心理学
政治学
心理学
标识
DOI:10.1115/dscc2020-3293
摘要
Abstract For safety-critical systems, the safety constraints are required to be satisfied through control design. To describe and apply system safety constraints, controlled invariant set (CIS) and control barrier function (CBF) are effective methods. In many applications, safety constraints may change dynamicly in operation environment and thus are not always continuous. However, such discontinuity and switching problems of CIS and CBF were well-addressed in the literature. In this paper, to tackle the issues, novel definitions of switched CIS (SCIS) and switched CBF (SCBF) are proposed. To address the undifferentiability issue in SCBF, a novel relaxation function that connects different SCISs is proposed. Sufficient conditions for selecting the relaxation function are proposed with a rigorous proof. In the application of vehicle safety control for an obstacle avoidance problem, a relaxation function is selected and demonstrated in the safety control design. To evaluate the effectiveness of the proposed definitions and control design, simulation results are presented and discussed.
科研通智能强力驱动
Strongly Powered by AbleSci AI