杂波环境下基于数据压缩的多传感器容积滤波算法

Algorithm of Muti-Sensor Cubature Filter Based on Data Compressionion Under Clutter Environment

摘要:

针对杂波环境下非线性系统中多传感器多目标跟踪问题,基于广义多维分配(S-D分配)规则获取最佳的量测划分,通过多传感器数据压缩技术得到等效量测点与等效量测协方差,结合容积卡尔曼滤波原理实现多目标跟踪,提出了一种基于数据压缩的多传感器容积滤波算法(SD-DCCKF)。仿真结果表明:相对已有算法,SD-DCCKF不仅避免了因模型线性化误差导致的滤波发散问题,而且克服了算法在高维系统中数值不稳定的缺点,算法估计精度较高,收敛速度较快,能够更加有效地解决非线性系统中的多目标跟踪问题。

Abstract:

According to the multi-sensor multi-target tracking problem of nonlinear systems in a cluttered environment, anovel cubature Klaman filter algorithm based on M-D assignment and data compressionion (MD-DCCKF) was proposed.In the new algorithm, the measurements from each sensor were permuted and combined, and the best combination wouldbe found according to a rule. Then, the best combination would be compacted into an equivalent measurement which wasassociated with targets by using the MSJPDA techniques. Finally, the targets’state would be estimated based on cubature Kalman filter. Simulation results showed that MD-DCCKF outperformed the existing algorithms in the aspects of trackingaccuracy, convergence rate and filtering robustness.