水下地磁导航适配区选取方法及其航迹规划

Selecting navigation matching region and planning optimal route are urgent issues to be settled in underwater vehicle geomagnetic navigation. Taking navigation matching area selection and route planning methods in underwater geomagnetic navigation as the research objects, this project carries out research in accordance with the following steps. Firstly, with applications of geomagnetic navigation information, earth magnetic field is analyzed, the function models of underwater geomagnetic navigation informative are obtained, then, correlation functions between underwater geomagnetic navigation error and geomagnetic features are constructed. Secondly, the impacts of geomagnetic feature informative parameters on matching algorithms’ calculating error are defined, optimal combination methods of geomagnetic parameters and matching algorithms are analyzed, then, systematic simulation tests are made by considering the following factors comprehensively, geomagnetic features, matching algorithms, and navigation accuracy. Then, matching area selection rules in underwater geomagnetic navigation are put forward. Finally, based on matching areas’ distribution, route planning evaluation models are established and verified with simulation, and optimal route is obtained. In theory, this project builds analysis models, including geomagnetic feature information analysis models, geomagnetic feature matching relational models, matrix models between geomagnetic matching algorithms and geomagnetic features, route planning evaluation models, etc. In application, the relevant algorithms are verified with geomagnetic data in some areas, and methods of navigation matching area selection and route optimization are proposed. This project is to clarify underwater vehicle geomagnetic navigation matching area selection methods and route planning optimization algorithms, and the final purpose is to improve geomagnetic navigation accuracy.

选取导航适配区，并规划基于适配区的最优航迹是水下航行器地磁导航应用中急需解决的难题。以水下地磁导航适配区选取及航迹规划方法为研究对象：首先，根据地磁场导航信息应用，解析地磁场，建立地磁场导航特征的函数模型，构造地磁特征信息量与地磁导航误差的相关函数；其次，明确地磁特征参数对匹配算法解算误差的影响，研究地磁参数与匹配算法的最优组合，并系统性地对地磁特征量、匹配算法、导航精度进行仿真测试，提出水下地磁导航适配区选取规则；最后，结合适配区分布特征，建立适配区航迹规划评估模型，并仿真验证，寻找最优地磁导航路径。通过本项目的研究，在理论层次上，建立地磁量信息分析模型、地磁导航特征匹配关联模型、地磁匹配算法与地磁特征矩阵关联模型、航迹代价及评估规划模型，等；在应用层次上，对相关算法进行实地数据验证，提出适配区选取及航迹规划方法。本研究旨在明确水下航行器地磁导航适配区及其优化航迹，提高地磁导航精度。

地磁匹配导航是实现水下航行器自主导航的重要导航方式。以地磁传感误差校正、适配区选取、匹配算法及航迹规划算法等关键技术为研究对象，建立地磁传感信息误差校正模型、地磁导航特征匹配关联模型，优化地磁匹配算法，提出匹配区参数/匹配因子等多约束条件下的航迹规划算法等，并对相关算法进行验证。. 针对随机高幅值磁干扰，提出一种基于多尺度干扰检测与完备集成经验模态分解（ID-CEEMD）相结合的降噪方法。根据地磁信号的强相关性，自适应地选择有效本征模态分量（IMF），并将有效IMF叠加进行信号重构。实验结果表明，提出的方法能有效滤除磁干扰，信噪比提高了3倍以上。. 针对匹配初始定位误差较大、搜索区域过大，传统的粒子初始化方法会出现粒子密度下降、算法收敛效率低的问题，提出一种约束有效定位的粒子群优化算法（CEPPSO）。算法通过对搜索区域置信概率密度函数的约束划定有效定位区，并在有效定位区内初始化粒子，提高粒子在真实值附近的概率和降低粒子数量；优化惯性权值更新，平衡了全局搜索与局部搜索能力。仿真结果表明，粒子初始化范围为常规搜索范围的1/3，同样粒子个数条件下，CEPPSO的匹配成功率比常规的PSO算法提高近两倍。. 分析基于遗传算法和支持向量机、分形维数、主成分分析法等适配区评价方法；提出地磁图特征参数的计算模型，分析各地磁图特征参数适配性的相互关联；建立多属性评价地磁图适配性的综合评价矩阵，采用基于Fuzzy集决策适配性的方法，并在此基础上提出了基于Vague集模糊决策的适配区选取规则。. 提出采用Dijkstra算法求解地磁图适配性约束下的航迹规划问题，实现地磁图适配性约束下的粗选航迹规划。对水下环境约束条件进行了分析与建模；将多约束条件进行了归一化处理，建立了航迹评价优化模型；对粒子群的惯性权重的选取以及粒子编码方式进行了改进；并采用Dijkstra算法+粒子群优化算法的混合航迹规划算法取得从起始节点要目标节点的最优航迹。实验结果表明多约束下的航迹规划算法比粒子群等算法更优。