基于仿生行为的自主水下机器人海洋溢油羽流追踪和测绘研究

For the requirements of autonomous underwater vehicles (AUVs) in the efficient survey of oceanic oil spill plumes, we propose in this project to do researches on on-line decision-making and planning methods and algorithms for AUVs to autonomously and efficiently track and mapp oil spill plumes, which are based on the approach and results of bio-inspired chemical plume tracing researches. In this project, following aspects are mainly to be investigated: (1)an effective simulation environment for the plume tracking and mapping research; (2)methods and algorithms for processing and fusion of the data from the in-situ plume sensors mounted on the AUV; (3)modeling and prediction of the oil spill plumes with real-time AUV's sensor data; (4)abstraction, design, optimization and validation of the AUV behaviors for the tracking and mapping of a oil spill plume with continuous source or without source. Through the researches, key problems related to the above researches will be resolved, and the research results will be demonstrated and validated via computer simulation and field experiments. By the research efforts of the project, we anticipate to obtain effective and efficient strategies, methods and algorithms which could enable an AUV to track and map oceanic oil spill plumes autonomously and efficiently. The research results will provide necessary therotical foundation and technical reserve for the future applications of AUVs in oceanic oil spill plume survey missions.

针对自主水下机器人（AUV）在高效的深海溢油羽流测绘中的应用需求，基于国际上基于仿生行为的自主机器人以"之"字形的运动策略追踪湍羽流的研究思路和结果，研究AUV基于探测传感器信息在线实时决策、规划，自主追踪溢油羽流并同时快速、高效的获取溢油羽流的位置、尺度、及其示踪物浓度时空分布等信息的方法和算法。主要围绕AUV溢油羽流追踪和测绘仿真研究环境，AUV溢油羽流原位探测传感器数据实时处理和融合，AUV溢油羽流在线建模及其时空分布预测，和AUV追踪和测绘连续源和无源溢油羽流的自主行为抽象、设计、优化、和验证等内容开展研究工作，研究并解决其中涉及的关键科学问题，并通过计算机仿真和外场实验对研究成果进行演示和验证。以期通过本项目的研究，获得AUV基于仿生行为高效的追踪和测绘海洋溢油羽流的策略、方法和算法，为将来我国AUV自主的快速、高效探测海洋溢油羽流的实际应用提供必要的理论基础和技术储备。