【作者】 刘杨；

【导师】 徐博； 杨羡；

【作者基本信息】 哈尔滨工程大学 ， 控制工程（专业学位）， 2015， 硕士

【摘要】 海洋开发是当前的热点问题,而认识和开发海洋需要各种高科技手段。自主式水下机器人(Autonomous Underwater Vehicle,AUV)作为其中一种重要手段,正日益受到重视。对AUV进行高精度的导航定位,是AUV完成各种工作任务的前提和保障,而协同导航正是一种对多AUV系统进行精确导航的有效方法。本文针对多AUV协同导航系统,对多AUV协同导航中的优化算法和编队构型问题展开研究。论文完成的主要工作有:1、首先以单领航艇多AUV协同导航系统为背景,详细介绍了多AUV协同导航的概念特征;其次描述了导航系统中常用的坐标系及其关系,定义了AUV常用运动参数;然后介绍了协同导航系统中所应用的传感器;最后介绍了多AUV协同导航的基本原理。2、基于状态空间法和AUV运动学方程,建立了单领航艇多AUV协同导航系统的数学模型,深入讨论了处理模型所适用的协同导航算法。研究了扩展卡尔曼滤波(EKF),超球体单形采样无迹卡尔曼滤波(SSUKF)和容积卡尔曼滤波(CKF),详细介绍了这三种滤波方法在单领航艇多AUV协同导航中的应用,并分别对其进行了分析和推导。最后,对上述协同导航算法进行了仿真验证,并从计算量和精度两个方面对其进行了分析比较。3、从可观测度分析与Cramer-Rao不等式下界极值分析两个角度,对多AUV协同导航系统的最优编队构型问题进行研究。给出了单领航艇情况下的可观测度和Cramer-Rao不等式下界计算方法,并根据计算结果分析设计了此时的最优编队构型。在此基础上,讨论了继续增加领航艇/跟随艇后的多AUV协同导航最优队形问题,计算了不同情况下的Cramer-Rao不等式下界,并根据计算结果进行了最优编队构型设计。4、对EKF、SSUKF和CKF三种多AUV协同导航算法,在单领航艇情况下分别设计各种队形的协同导航仿真实验,通过分析实验结果,对三种算法和它们在不同队形下的导航精度进行了讨论;然后,继续增加领航艇的数量,进行多领航艇不同队形下的协同导航仿真实验,比较协同导航精度,从而证明所设计的最优队形的正确性。最后,进行了双领航艇多AUV协同导航的实际实验,通过对实验数据进行分析,验证了队形与导航精度的关系和所设计的最优编队构型。更多还原

【Abstract】 The development of ocean is a hotspot nowadays,and it needs many kinds of high-tech methods to understand and develop the ocean.As one of the important methods,Autonomous Underwater Vehicle(AUV)has been paid increasing attentions nowadays.High precision navigation and positioning of AUVs is the premise and guarantee for AUVs to complete various tasks,and cooperative navigation is an effective method of accurate navigation for multi-AUV system.In this paper,aiming at multi-AUV cooperative navigation system,the optimization algorithm and formation configuration have been studied.The main contents are as follows.1.Firstly,the concept and features of multi-AUV cooperative navigation is introduced in detail under the background of single master multi-AUV cooperative navigation system.Then the coordinate systems which are commonly used in navigation system and their relationships are described,and the commonly used AUV motion parameters are defined.Then the sensors used in cooperative navigation system have been introduced in details.Finally,the fundamental of multi-AUV cooperative navigation system is introduced.2.Based on the state-space method and kinematical equation of AUV,the mathematical model of single master multi-AUV cooperative navigation system is built,and cooperative navigation algorithms which are fit for the processing of the model have been discussed deeply.The Extended Kalman Filtering(EKF),the Simplex Sampling Unscented Kalman Filtering(SSUKF)and the Cubature Kalman Filtering(CKF)have been introduced detailedly,and their analysis and derivation have been proceeded,respectively,then simulation of the application of above algorithms in single master multi-AUV cooperative navigation system have been carried out compared by their calculation amount and accuracy.3.Based on the single master multi-AUV cooperative navigation system mode,as well as extend it into more masters mode,study the optimized formation configuration of multi-AUV cooperative navigation system on basis of two theories: the observability and the extreme value of Cramer-Rao inequality lower bound,the optimal formation has been analyzed,and the optimized formation of single master,double masters,adding more master/slave AUVs have been studied,the value of Cramer-Rao inequality lower bound is calculated,and finally come up with the design of multi-AUVs cooperative navigation system’s optimal topology formation configuration under different conditions.4.For the above mentioned three single mater multi-AUV cooperative navigation algorithms together: EKF,SSUKF and CKF,the single master cooperative navigation simulation experiments has been designed,respectively.By analyzing the experimental results,the characteristics in navigation accuracy and system stability of three kinds of algorithms are discussed.Then by increasing the number of master AUV,the simulation test and verify of the cooperative navigation accuracy have been made under different formation to prove the correctness of the designed optimized formation.At last,an actual experiment of double master multi-AUV cooperative navigation has been carried out,and by analyzing the experimental data,the relationship of formation and navigation accuracy is verified,and the designed optimal formation configurations is proved.更多还原