| 引用本文: | 岳伟,辛弘,林彬,刘中常,李莉莉.MAUV协同搜索多智能目标的路径规划[J].控制理论与应用,2022,39(11):2065~2073.[点击复制] |
| YUE Wei,XIN Hong,LIN Bin,LIU Zhong-chang,LI Li-li.Path planning of MAUV cooperative search for multi-intelligent targets[J].Control Theory and Technology,2022,39(11):2065~2073.[点击复制] |
|
| MAUV协同搜索多智能目标的路径规划 |
| Path planning of MAUV cooperative search for multi-intelligent targets |
| 摘要点击 1030 全文点击 223 投稿时间：2021-11-22 修订日期：2022-02-07 |
| 查看全文 查看/发表评论 下载PDF阅读器 |
| DOI编号 10.7641/CTA.2022.11140 |
| 2022,39(11):2065-2073 |
| 中文关键词 多自主式水下机器人协同搜索 马尔科夫链 目标概率图 改进的多狼群算法 |
| 英文关键词 multi-AUV cooperative search Markov chain target probability map improved multi-wolf pack algorithm |
| 基金项目 大连市科技创新基金项目(2019J12GX040), 中央高校基本科研业务费项目(3132019355), 大连高层次人才创新支撑计划项目(2019RQ057, 2020RQ060)资助. |
|
| 中文摘要 |
| 本文针对复杂水下环境中多自主式水下机器人(MAUV)协同搜索多个智能目标这一重要课题展开研究. 首
先, 利用马尔科夫链建立智能目标的决策状态转移模型的同时考虑了智能目标决策与行动的对应关系, 并结合不同
光照反射强度下传感器探测概率受限模型, 设计新的目标概率图更新策略. 然后, 结合MAUV系统的约束条件和搜
索效率建立实时适应值函数. 接着, 本文提出一种改进的多狼群算法(IMWPA)搜索策略, 包括: 1) 利用人工势场法
调整步长因子, 使启发式算法更加适应探索过程. 2) 设计多狼群嚎叫环节, 建立了狼群间的信息交流渠道. 3) 提出
新的狼群淘汰更新机制, 保障了人工狼多样性的同时避免算法趋于完全随机. 最后, 通过MATLAB仿真实验对比验
证了本文算法的可行性及优越性. |
| 英文摘要 |
| This paper focuses on the important issue of collaborative search for multi-intelligent targets by multiautonomous
underwater vehicle (MAUV) in complex underwater environment. Firstly, the Markov chain is used to establish
the decision state transition model of the intelligent target. Considering that the intelligent target takes different
actions under different decision states, and combining with the sensor detection probability constraint model under different
light reflection intensity, a new target probability map update strategy is designed. Combined with the constraint
conditions and search efficiency of MAUV system, the real-time fitness function is established. Then, an improved multiwolf
pack algorithm (IMWPA) search strategy is proposed. 1) The artificial potential field is used to adjust the step size
factor, so that the heuristic algorithm is more suitable for the exploration process. 2) Designing multi-wolf howl link to
establish a channel for information exchange between wolves. 3) A new wolf elimination update mechanism is proposed to
ensure the diversity of artificial wolves and avoid the algorithm becoming completely random. Finally, the feasibility and
superiority of the proposed algorithm are verified by MATLAB simulation. |
|
|
|
|
|