仿虾偏振信息计算及水下机器人导航定位方法研究

New type technology of the underwater navigation is the interest topic over the word in this time. With the increasing demand for underwater applications in diverse environments i.e. the rivers, lakes and offshore/onshore, the autonomous underwater robots have increasingly applications in the underwater rescue, underwater structure damage detection and the protection of the coastal engineering. However, the water situation is complicated in the river, lake and offshore/onshore in our nation. The multi-phase flow, high sediment, high algae, erosion and deposition interact would make the underwater landform significantly change. This nature highlights the drawback of deep-sea navigation and positioning technology when it is used in the the river, lake and offshore/onshore environments. Aiming to handle this issue, this project intents to break the intrinsic limitation of the existing navigation technologies, i.e. acoustic, gravity, geomagnetic, inertial and polarization based method, when they are introduced to complicated and variable underwater environments. We will explore the distribution pattern of the polarization E-vector and navigation mechanism under water. Also this pattern is connected to the polarization sensitive mechanism of the mantis shrimp which is imitated and modelled. Imitating to the adaptive polarization sensitivity that is combined to the underwater polarization E-vector analysis, the heading and position measurement method of the will be developed. The results of this research will significantly promote the development of the new type of underwater navigation and positioning technology.

新型水下导航技术是目前研究的国际前沿和热点。随着对河流湖泊、近海/岸水下探测应用需求日益增加，可自主运行并能承担探测任务的水下机器人在水下救助、水下构筑物损毁检测、海岸工程防护等方面应用日趋广泛。然而，我国内陆河湖及近海水况复杂，多相流、高泥沙和高藻类水体，及冲刷、淤积造成的水下地貌动态变化，使得面向深海水体环境的水下导航定位技术直接移植应用于内陆河湖及近海/岸存在诸多不足。为此，本项目试图突破现有的水声/重力/地磁/惯性/偏振等导航技术如何增强应用于水下复杂多变性探测环境的适应能力所遇到的本质性困难，通过揭示水下偏振E矢量方向分布模式及水下偏振导航机理、模拟水下生物-螳螂虾复眼视觉系统的功能优势、仿虾偏振敏感机制的适应性信息获取、水下偏振E矢量解译及航向与位置测算等探索研究，提出仿虾偏振信息计算及水下机器人导航定位新方法。成果将有助于推进新型水下导航定位技术的发展。

我国内陆河湖和近海/岸水流水况较为复杂，水下环境常面临多相流、高泥沙和高藻类等情况，冲刷、淤积造成水下地貌动态变化，对水下机器人自主式导航定位产生较大影响。水下机器人自主导航定位的重点是对航向、位置信息的识别。然而，现有水声/重力/地磁/惯性等导航定位技术在诸如内陆河湖水库及近海/岸水下探测等实际应用中却存在众多不适应。.本课题基于偏振导航机理，围绕水下偏振分布模式特性和仿虾偏振计算开展水下机器人偏振导航方法研究，借鉴生物偏振导航机理和螳螂虾偏振视觉机制研究成果，从偏振E矢量分布模式和仿生偏振信息处理研究入手，重点研究和探索了偏振信息获取方式、偏振信息计算方式、水下机器人导航定位算法等，包括：（1）研究了生物偏振导航机理、偏振模式变化规律，建立了基于水下偏振E矢量方向分布与太阳位置间对应关系的水下偏振导航研究策略和技术框架。（2）研究了螳螂虾偏振敏感机理、视域偏振特征表征的自适应处理机理、虾视觉系统正交侧抑制信息处理模式，提出了仿虾复眼正交对立的偏振信息敏感机制的视域场景特征表示、不同偏振方向上的强度特征及互相关特性、仿虾视觉水下场景偏振表征E矢量等方法，其比传统偏振信息计算方法更具优势。（3）提出了基于偏振特性的水下定位方法，水下机器人位姿计算与优化、水下复杂环境下全局路径规划等算法，导航定位效能优于现有方法。最后，研发了一种仿虾偏振信息获取装置。.本课题探索转化螳螂虾视觉系统对偏振计算和偏振导航定位的途径，设计的研究策略、建立的技术框架、提出的创新方法能有效克服现有水下偏振计算在建模、计算和导航定位等方面的困难。构建了新型水下机器人偏振导航定位的关键理论与方法，为有效解决水下机器人导航问题提供了新技术途径。.课题执行期间发表论文12篇，申请国家发明专利9项，培养硕士研究生5名。