滤波初值对测向定位的影响及其消除方法

Influence and Elimination of Algorithm Initial Value in Bearing-Only Passive Location

  • 摘要: 基于运动单站测向交叉定位中将滤波初值分成两类,该文从几何原理上解释了不同类型的滤波初值对定位结果的影响;提出了通过增加相位差观测量,利用相位差的变化率和方位角信息,将非线性系统转换成线性系统,再利用卡尔曼滤波来改善定位结果对滤波初值依赖性的方法。转换后的线性系统在利用卡尔曼滤波时,初始值可以根据以前的测量值得到。模拟仿真表明该方法能有效摆脱定位结果对滤波初值的依赖,且定位精度在2%以内。

     

    Abstract: The influence of two kinds of algorithm initial values on results in bearing-only passive location is described with geometry method. Via adding the difference of phase as a new measurement vector, converting the nonlinear system to linear system, a way to avoid this influence is proposed. In the linear system, Kalman filtering (KF) is used, and the initial values can be gotten from the measured values. Simulations show that this way can eliminate the initial value's influence, and the relative error of distance is less than 2%.

     

/

返回文章
返回