基于多信标的深水机器人导航与同时定位方法研究
文献类型：学位论文
| 作者 | 王飞
|
| 学位类别 | 硕士 |
| 答辩日期 | 2013-05-28 |
| 授予单位 | 中国科学院沈阳自动化研究所 |
| 授予地点 | 北京 |
| 导师 | 刘健 |
| 关键词 | 自主水下机器人 长基线系统 扩展卡尔曼滤波 M估计 同时定位 |
| 其他题名 | Multi-beacon Based Simultaneous Localizationand Mapping of Deep Water Robot |
| 学位专业 | 机械电子工程 |
| 中文摘要 | 水下机器人航行中的导航定位问题是水下机器人研究领域最基本的问题之一。基于测距声信标的长基线定位技术，是目前水下机器人导航中最为经典和广泛使用的手段。它不仅具备很高的局部定位精度，安全性和可靠性较好，而且可以降低水下机器人的导航传感器成本，从而推动了水下机器人技术的发展和应用。然而实际中长基线系统存在布防繁琐、基阵作用范围有限等问题，这也在一定程度上限制了它自身的发展。鉴于水下机器人的发展态势和长基线系统在水下机器人导航应用领域的重要地位，需要对长基线导航算法进行深入的研究，并结合实际需要设计出相应的组合导航算法。 本文的研究工作以“6000米无人无缆潜器实用化改造”项目为依托，在总结国内外研究成果的基础上，借鉴其他领域的成果，重点研究了长基线M-EKF导航算法和基于多信标的AUV同时定位算法。具体的研究内容为： （1）长基线系统仿真 本部分对长基线系统仿真模块的建立过程进行了介绍。首先介绍了长基线系统工作的原理；然后介绍了水下射线声学的基本理论，并讨论了声速剖面数据的处理方法；最后结合实际原理，给出了长基线系统的数学仿真模型，并通过二分法搜索实现了模型仿真。这些工作为长基线导航方法的研究工作提供了仿真环境支持。 （2）长基线导航算法研究 本部分首先设计并讨论了6000米深水机器人EKF导航算法，并依据仿真结果论证了该算法的优缺点；然后针对算法的鲁棒性问题，设计了基于M-EKF的深水机器人导航方法，通过向原算法中引入M估计，达到抑制数据异常值的目的，仿真试验证明了该方法的有效性。 （3）基于多信标的AUV同时定位算法 由于长基线系统布防繁琐、基阵作用范围有限，该部分提出了基于多信标的深水机器人SLAM算法。首先对SLAM模型中的各个分模型依次进行了讨论；然后针对SLAM模型中的一些不确定性，设计了声信标观测异常值剔除方法和多信标的构图方法；最后构造了系统的状态和观测方程，并在EKF的基础上采用状态扩展方法，进而实现了同时定位功能。 |
| 英文摘要 | The underwater robot navigation is one of the most fundamental problem of the field of underwater robotics research. Acoustic ranging-based long baseline positioning technology is the most classic and widely used means of underwater robot navigation. Not only does it have a high local positioning accuracy, better security and reliability, but also it can reduce the cost of underwater robot navigation sensors, and then promote the development and application of underwater robots. The long baseline system, however, has some practical issues such as cockamamie laying process, limited coverage of arrays, which restrict its development in a way. In view of the development trend of the underwater robot and the important position of long baseline system in the field of underwater robot navigation applications, it is necessary to deeply research long baseline navigation algorithm and to design integrated navigation algorithm according to actual requirement. The work in this paper is based on the item of “6000 meters autonomous underwater robot practical transformation”. On the basis of summing up the research results at home and abroad, a long baseline M-EKF navigation algorithm and a multi-beacon based AUV simultaneous localization algorithm are proposed. The specific research content: (1)Long Baseline System Simulation This section gives the introduction of establishing long baseline system simulation module. Firstly, the principle of long baseline system is presented. Secondly, the basic theory of underwater acoustic is introduced, and then the way of dealing with sound velocity profile is discussed. Lastly, the mathematical model of long baseline system simulation is proposed based on the actual principle, here, binary search is adopted for realizing the module. All these work above gives an important simulation environment support for the research of long baseline navigation algorithm. (2)Long baseline navigation algorithm research This section first designs and discusses underwater robot EKF navigation algorithm, then the advantages and disadvantages of the algorithm are demonstrated based on the simulation results; To improve the robustness of the EKF navigation algorithm, an M-EKF navigation algorithm is proposed, by the use of M-estimate method in EKF algorithm, the system outliers can be restrained effectively, the simulation result has demonstrated its effectiveness. (3)Simultaneous localization algorithm of AUV based on multi-beacon Because long baseline system’s arming is cumbersome, and it’s coverage is limited, this section gives the deep-water robot SLAM algorithm based on multi-beacon. Firstly, each part of the SLAM model is discussed. Secondly, to resolve some uncertainty in the SLAM system, a method for beacon measurement outlier rejection and an approach of mapping for multi-beacon are proposed. Lastly, the state equation and the observation equation of the system are constructed, then, one way of state extension is introduced in the EKF algorithm, which realizes the function of simultaneous localization. |
| 语种 | 中文 |
| 公开日期 | 2013-08-19 |
| 产权排序 | 1 |
| 页码 | 59页 |
| 分类号 | TP242 |
| 源URL | [http://ir.sia.ac.cn/handle/173321/10736]  |
| 专题 | 沈阳自动化研究所_海洋信息技术装备中心 |
| 推荐引用方式 GB/T 7714 | 王飞. 基于多信标的深水机器人导航与同时定位方法研究[D]. 北京. 中国科学院沈阳自动化研究所. 2013. |
入库方式： OAI收割
来源：沈阳自动化研究所
浏览0
下载0
收藏0
其他版本
除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。