改进重采样的高斯粒子滤波定位技术

Wireless Location for Gaussian Particle Filter of Improved Re-Sampling

针对粒子滤波算法中仍然存在的样本匮乏现象,在高斯滤波的基础上,提出了一种改进重采样的高斯粒子滤波(IR-GHPF)算法。经过新的重采样后的粒子包含了更多相邻粒子的状态信息,提高了粒子的多样性。将此算法应用于无线定位系统中,仿真结果表明,该算法在NLOS环境下仍然具有较高的估计精度,其定位性能优于粒子滤波算法和高斯粒子滤波算法。

In order to reduce the phenomenon of sample impoverishment in particle filter,an improved re-sampling of Gaussian particle filter（IR-GHPF） algorithm is proposed based on the Gaussian filtering.The particles after new resampled,contain more state information of adjacent particles and enhance the diversity of particles.The algorithm is applied to the wireless positioning system,the simulation results show that IR-GHPF has high estimation precision in the NLOS environment and the positioning performance is superior to particle filter algorithm and Gaussian particle filter algorithm.