在大范围海图数据环境中,应用遗传算法(GA)对自主式水下潜器(简称 AUV)的全局路径规划问题进行了研究,介绍了基于栅格的环境模型及其数据结构,讨论了 GA 的染色体编码方式、基于知识的初始种群生成方法与适应度函数,基于领域知识设计了...在大范围海图数据环境中,应用遗传算法(GA)对自主式水下潜器(简称 AUV)的全局路径规划问题进行了研究,介绍了基于栅格的环境模型及其数据结构,讨论了 GA 的染色体编码方式、基于知识的初始种群生成方法与适应度函数,基于领域知识设计了五种遗传算子。通过仿真结果可以看出:采用可变长编码方式使路径描述简单、清晰,算法具有收敛速度快、求解实际问题效率高的特点。展开更多
Path planning is an important issue for autonomous underwater vehicles （AUVs） traversing an unknown environment such as a sea floor, a jungle, or the outer celestial planets. For this paper, global path planning usi...Path planning is an important issue for autonomous underwater vehicles （AUVs） traversing an unknown environment such as a sea floor, a jungle, or the outer celestial planets. For this paper, global path planning using large-scale chart data was studied, and the principles of ant colony optimization （ACO） were applied. This paper introduced the idea of a visibility graph based on the grid workspace model. It also brought a series of pheromone updating rules for the ACO planning algorithm. The operational steps of the ACO algorithm are proposed as a model for a global path planning method for AUV. To mimic the process of smoothing a planned path, a cutting operator and an insertion-point operator were designed. Simulation results demonstrated that the ACO algorithm is suitable for global path planning. The system has many advantages, including that the operating path of the AUV can be quickly optimized, and it is shorter, safer, and smoother. The prototype system successfully demonstrated the feasibility of the concept, proving it can be applied to surveys of unstructured unmanned environments.展开更多
文摘在大范围海图数据环境中,应用遗传算法(GA)对自主式水下潜器(简称 AUV)的全局路径规划问题进行了研究,介绍了基于栅格的环境模型及其数据结构,讨论了 GA 的染色体编码方式、基于知识的初始种群生成方法与适应度函数,基于领域知识设计了五种遗传算子。通过仿真结果可以看出:采用可变长编码方式使路径描述简单、清晰,算法具有收敛速度快、求解实际问题效率高的特点。
基金Supported by State Key Laboratory of Robotics and System (HIT) under Grant No.SKLRS200706the Heilongjiang Scientific Research Foundation for Postdoctoral Financial Assistance under Grant No.323630221the Project of Harbin Technological Talent Research Foundation under Grant No.RC2006QN009015
文摘Path planning is an important issue for autonomous underwater vehicles （AUVs） traversing an unknown environment such as a sea floor, a jungle, or the outer celestial planets. For this paper, global path planning using large-scale chart data was studied, and the principles of ant colony optimization （ACO） were applied. This paper introduced the idea of a visibility graph based on the grid workspace model. It also brought a series of pheromone updating rules for the ACO planning algorithm. The operational steps of the ACO algorithm are proposed as a model for a global path planning method for AUV. To mimic the process of smoothing a planned path, a cutting operator and an insertion-point operator were designed. Simulation results demonstrated that the ACO algorithm is suitable for global path planning. The system has many advantages, including that the operating path of the AUV can be quickly optimized, and it is shorter, safer, and smoother. The prototype system successfully demonstrated the feasibility of the concept, proving it can be applied to surveys of unstructured unmanned environments.
