不确定海流环境下水下机器人最优时间路径规划
Time-optimal path planning for autonomous underwater vehicles with uncertain ocean currents
摘要点击 64  全文点击 64  投稿时间:2019-04-20  修订日期:2019-11-11
查看全文  查看/发表评论  下载PDF阅读器
DOI编号  10.7641/CTA.2019.90273
  2020,37(6):1302-1310
中文关键词  水下机器人  路径规划  不确定性  区间优化
英文关键词  autonomous underwater vehicles  path planning  uncertainty  interval optimization
基金项目  中国高技术船舶科研项目(KY10400170181)资助.
作者单位邮编
姚绪梁 哈尔滨工程大学 自动化学院 150001
王峰 哈尔滨工程大学 自动化学院 
王景芳 哈尔滨工程大学 自动化学院 150001
王晓伟 江西九江职业技术学院 
中文摘要
      为解决海流预测不精确条件下, 现有基于确定性海流路径规划算法鲁棒性差和规划的路径有可能为不可 行路径的问题, 本文提出一种基于区间优化的水下机器人(AUV)最优时间路径规划算法. 该算法采用双层架构, 外 层用蚁群系统算法(ACS)寻找由起点至终点的候选路径; 内层以区间海流为环境模型, 计算候选路径航行时间上下 限, 并分别通过区间序关系和基于可靠性的区间可能度模型将航行时间区间转换为确定性评价函数, 并将评价函数 值作为候选路径适应度值返回到外层算法. 仿真结果表明, 相对于确定海流场路径规划方案, 提出的方案增强了路 径规划器的鲁棒性并解决了结果路径不可行问题.
英文摘要
      Under the condition of inaccurate measurement or prediction of ocean currents, the existing path planning algorithms based on deterministic ocean currents have low robustness and the resultant paths may be infeasible. To solve these problems, this paper presents a time-optimal path planning algorithm for autonomous underwater vehicles (AUVs) based on the interval optimization. The proposed scheme adopts a two-layer architecture. The outer layer uses ant colony system (ACS) algorithm to find candidate path from starting point to end point; In the inner layer, the upper and lower bounds of travelling time for candidate path are calculated under the environment model of interval ocean currents. Then, the travelling time interval is transformed into a deterministic value by interval order relation and reliability-based possibility degree of interval, and the value is returned to the outer layer as the fitness value of candidate path. The simulation results show that, comparing with the path planning scheme with deterministic ocean currents, the proposed scheme enhances the robustness of path planner and solves the problem that the resultant path may be infeasible.