基于遗传算法的非线性最小二乘在纯方位系统中的应用

Application of nonlinear least square based on genetic algorithms in bearings-only target motion analysis

鉴于非线性最小二乘法在国内目标运动分析中应用较少的现状,研究了纯方位系统的非线性最小二乘解法,提出了遗传算法的求解方法。结合实际问题设计了具体的遗传算法,并用实例进行了仿真分析。通过仿真结果的比较,可以看出用遗传算法的非线性最小二乘法明显优于线性方法,也优于非线性方法的梯度解法。

As nonlinear least square（NLS） is rarely applied in target motion analysis（TMA）interiorly, NLS of bearings-only TMA is studied and methods to solve NLS based on genetic algorithms（GA） are put forward. According to practice of problems,GA is designed in detail,and simulations with analysis are given. After comparing the result of simulation,it is concluded that NLS based on GA is superior to linear square in evidence,also to NLS based on gradient algorithms.