摘要
针对水下被动目标跟踪的非高斯噪声环境和弱可观性的特点,提出了将粒子滤波算法应用于水下被动目标跟踪的思路。该算法直接利用传感器获得的含有噪声的角度数据,通过改进极坐标系下的系统方程得到目标状态的后验概率分布,来估计目标的运动状态。仿真结果表明该算法提高了滤波的稳定性,跟踪精度优于扩展卡尔曼滤波算法和无迹卡尔曼滤波算法。
The method base on PF for motion estimation is proposed according to the characteristics of non-Gaussian noise environments and weak observability in underwater passive target tracking system.The algorithm utilizes angle data of sensor with noise.Posteriori probability distribution of target is obtained through system equation in Modified Polar Coordinates and the target state is estimated. The simulation results show that the algorithm enhance the stability of filter and has good performances of tracking accuracy than extended Kalman filter and unscented Kalman filter.
出处
《电子设计工程》
2014年第8期74-76,80,共4页
Electronic Design Engineering
关键词
纯方位跟踪
贝叶斯估计
扩展卡尔曼滤波
粒子滤波
bearing-only tracking
Bayesian estimation
extended Kalman filter
particle filter
