基于自适应粒子滤波算法的对海跟踪研究

A Study on Ship Tracking Based on Adaptive Particle Filter

机载雷达对海面慢速运动目标跟踪时存在非高斯非线性状态估计问题,传统的卡尔曼滤波器跟踪不仅会产生目标定位偏差,而且会造成航迹起伏。文中将自适应粒子滤波算法应用到具有闪烁噪声背景下的海面目标跟踪问题中,此算法可根据预测粒子在状态空间中的分布情况自适应选择粒子数量,从而在保证跟踪精度的同时减少了算法的运算量。仿真结果表明,自适应粒子滤波器可实现对闪烁噪声背景下的慢速目标高精度定位跟踪,且跟踪性能优于标准粒子滤波器,具有工程实用价值。

Locating remote slow sea-surface targets with airborne radar is a problem of the non-linear and non-Gaussian state estima- tion, resulting in the deviation of the target location and the track jitter when track jitter dealing with traditional Kalman filter. In this paper, adaptive particle filter （PF） is used to solve the ship tracking problem under glint noise. The algorithm can choose the number of particles based on KLD-sampling, which is efficient and has smaller time consumption. Computer simulation proves that adaptive particle filter can quickly realize slow target tracking under glint noise, and the tracking performance is superior to that of standard PF and that adaptive PF is practically valuable for engineering use.