航空学报 > 2022, Vol. 43 Issue (10): 527129-527129   doi: 10.7527/S1000-6893.2022.27129

无人系统生存智能与安全、免疫、绿色控制技术

郭雷1,2, 朱玉凯2,3, 乔建忠1,2, 郭康3,4, 包为民2,3,4   

  1. 1. 北京航空航天大学 自动化科学与电气工程学院, 北京 100083;
    2. 鹏城实验室 数学与理论部, 深圳 518055;
    3. 北京航空航天大学 宇航学院, 北京 100083;
    4. 中国航天科技集团有限公司 科技委, 北京 100048
  • 收稿日期:2022-03-08 修回日期:2022-03-17 发布日期:2022-05-09
  • 通讯作者: 包为民,E-mail:baoweimin@cashq.ac.cn E-mail:baoweimin@cashq.ac.cn
  • 基金资助:
    国家自然科学基金(61127007,61627810,62103013,62122007);博士后创新人才支持计划(BX20200031)

Survival intelligence and safety, immunity and green control technologies for unmanned systems

GUO Lei1,2, ZHU Yukai2,3, QIAO Jianzhong1,2, GUO Kang3,4, BAO Weimin2,3,4   

  1. 1. School of Automation Science and Electrical Engineering, Beihang University, Beijing 100083, China;
    2. Department of Mathematics and Theories, Pengcheng Laboratory, Shenzhen 518055, China;
    3. School of Astronautics, Beihang University, Beijing 100083, China;
    4. Science and Technology Commission, China Aerospace Science and Technology Corporation, Beijing 100048, China
  • Received:2022-03-08 Revised:2022-03-17 Published:2022-05-09
  • Supported by:
    National Natural Science Foundation of China (61127007, 61627810, 62103013, 62122007); China National Postdoctoral Program for Innovative Talents (BX20200031)

摘要: 物竞天择,适者生存。动物经过亿万年进化,面对病毒侵袭、环境剧变、天敌侵害、种群竞争等挑战,具备了极端环境下的生存和繁衍能力。与动物一样,无人系统同样面临干扰、攻击、拒止、损伤、故障等不确定和异常因素的影响。在干扰对抗态势下,保证无人系统是"会学习"还是"能生存",已成为一个挑战性问题。本文从控制论的角度提出无人系统的生存智能问题以及3个相关的关键控制要素:安全控制、免疫控制和绿色控制。其中,安全控制涉及无人系统对于多源干扰和故障的抗扰与容错控制;免疫控制涉及无人系统对于对抗与竞争态势的自主感知、自我调节以及学习进化问题;绿色控制涉及无人系统在多约束下的"节能" "节时" "省力"和"省心"等控制问题。目标是使无人系统从智能行为和功能的角度具备"赋生"(视系统如生命、器件似器官)能力,实现从"方法论""系统论"设计到"行为论"设计的跨域,提升无人系统在危险、极端、特殊、恶劣环境下的生存能力。

关键词: 无人系统, 生存智能, 安全控制, 免疫控制, 绿色控制, 干扰消纳

Abstract: Natural selection makes survival of the fittest. During billions of years of evolution, animals have formed the survival and reproduction abilities in the extreme environment in the presence of various challenges (e.g., virus invasion, environmental upheaval, natural enemy invasion, and population competition). Similar to animals, unmanned systems are affected by kinds of uncertain and abnormal factors including disturbance, attack, denies, damage, and fault. In the disturbance and confrontation environments, enabling unmanned systems to "self-learning" or "survival" has become one challenging problem. From the perspective of control, this paper proposes the concept of survival intelligence and three related key control factors of unmanned systems:safety control, immunity control, and green control. Safety control refers to the anti-disturbance and fault-tolerant control of unmanned systems subject to multiple disturbances and faults. Immunity control includes the autonomous sensing, self-regulation, and learning evolution of the system in face of confrontation and competition. Green control implies the control for "energy-saving""time-saving""strength-saving" and "burden-saving" of the unmanned systems under multiple constraints. Systems are like life, and devices are like organs. The purpose is to give life to unmanned systems in terms of intelligent behaviors and functions, and to achieve the leap from "method design""system design" to "behavior design", so as to improve the survival ability of unmanned systems in dangerous, extreme, special, and harsh environments.

Key words: unmanned system, survival intelligence, safety control, immunity control, green control, disturbance absorption

中图分类号: