【作者】 刘予学；

【导师】 边信黔；

【作者基本信息】 哈尔滨工程大学 ， 控制理论与控制工程， 2003， 硕士

【摘要】 在自治式水下机器人(AUV)分层递阶模式的自主控制体系结构中，使命规划属于最高层的控制。为了实现AUV在复杂的非线性和非结构海洋环境中自主规划、远程航海的能力，并且具有高可靠性与强鲁棒性，使命规划是一项关键技术。以“海洋地貌勘测”为一个具体案例，本文描述了根据使命信息、海区海图、以及AUV自身状态进行使命路径规划，并在全局路径的基础上进行使命分解即任务描述的过程。 本文描述了采用基于栅格类环境描述的激活值传播算法进行海洋环境下的路径规划，并给出了无海流情况下规划所得路径的仿真曲线。在论述了多任务、实时操作系统QNX的结构特点以及基于QNX与Windows运行的PC机之间网络进程通信的基础上，本文描述了在QNX上开发的使命规划软件的结构功能和流程图，并叙述了仿真联调实验的过程，动态地显示了“海洋地貌勘测”这一案例使命规划的结果，并证明了使命规划所得的任务序列在实际运行中逻辑的正确性与可行性。更多还原

【Abstract】 The mission planning is the highest-lever control in the hybrid control architecture of autonomous underwater vehicle(AUV). The mission planning plays the important role in the AUV’s control system to implement the AUV’s self-planning and long-distance navigation capability in the nonlinear and non-structure ocean environment, with higher reliability and robustness. Taking the "ocean physiognomy reconnaissance" as a specific case , this paper describes the global path-planning based on mission information, specific chart, and AUV’s self-state, and the procedure of mission decomposition, named task descript in the foundation of the produced path.The paper describes the path planning in the ocean environment using activity propagation arithmetic based on grid environment presentation, and gives the results of path-planning in the case of no ocean flow. After discuss the structure and character of operating system QNX and inter-process communication between PC’s running QNX or Windows, the paper describes the structure, function and flow chart of mission planning software which is developed in QNX, and narrates the course of simulation co-debug experiment , dynamically showing the results of the mission planning in the case of "ocean physiognomy reconnaissance", and proving the logical correctness and feasibility of task serial produced by mission planning.更多还原