中文摘要：

针对现有地磁场匹配算法对载体沿直线前进假设的不足,提出可适用于非直线形路径的地磁场匹配改进方法。将匹配定位视作多级优化问题,采用多级蚁群优化算法。在运算的每一级利用距离度量,路径序列比对采用相似性度量的方式,实现蚁群算法的两重适应度函数计算,并通过实验仿真讨论了权值参数与测量序列长度对路径匹配结果的影响。仿真结果表明,该方法可以仅利用磁测数据,以适当的计算量,有效实现任意形状路径的匹配。

英文摘要：

The existed Magnetic Field Contour Matching（MAGCOM） algorithms based on correlation measure suppose that the vehicle travel in a straight path,thus we proposed an improved method which is applicable for nonlinear courses.The multi-level ant colony optimization with multiple fitness functions was adopted taking the geomagnetic locating as a multi-level optimization problem.The fitness functions calculated distance in each level and similarity in the whole sequence.The influences of the weight coefficient and length of sequence on path matching result were analyzed by experiments.Simulation results show that this approach can implement matching of arbitrary form paths effectively with moderate amount of computation by using only the measured magnetic data.