针对低信噪比微弱目标的实时检测与跟踪问题，提出了基于复似然比的粒子滤波改进算法。该算法弥补了传统粒子滤波利用幅度似然比计算粒子权重而忽略相位信息的缺陷，推导了基于复似然比的粒子权重表达式，从而更好地利用了目标原始信息。同时，基于幅度似然比的权重计算需要多次进行贝塞尔函数运算，而基于复似然比的权重计算只需进行一次贝塞尔函数运算，可以有效降低计算复杂度。仿真结果表明：改进算法不仅在检测与跟踪性能上优于传统粒子滤波算法，所需计算时间也明显降低。

An improved particle filter algorithm based on complex likelihood ratio is presented for weak target detection and tracking with low signal to noise ratio. The particle weight is achieved by real likelihood ratio without considering the phase information in the standard particle filter algorithm, which can be compensated by using the proposed algorithm and the original information is used more effectively. On the other hand, the operation of Bessel function is needed many times when calculating the particle weight based on the real likelihood ratio, while that is demanded only once to calculate the particle weight based on the complex likelihood ratio. Therefore, the computational complexity can be reduced by the proposed algorithm effectively. Simulation results show that the proposed algorithm has not only an improved performance of detection and tracking, but also costs much less computation time compared with the standard particle filter algorithm.

TN957.52

国家自然科学基金资助项目