中文摘要：

针对水下小目标粒子滤波估计过程中“粒子贫化”引起的估计性能下降，提出了混合粒子滤波算法。该算法在常规粒子滤波算法基础上，在每一步迭代估计过程中进行量测的再次随机采样，以丰富随机粒子多样性，缓解水下小目标状态估计过程中的“粒子贫化”的影响。对算法进行了仿真分析，并将该方法用于水下小目标探测实验的数据处理。结果表明，相比于常规的粒子滤波算法，所提出的混合粒子滤波得到了误差更小且稳定的状态估计结果，有效地改善水下小目标跟踪的精度和稳健性。

英文摘要：

In order to solve the problem of estimation performance degradation caused by particle impoverishment for underwater small target tracking, the mixture particle filter is proposed. The target measurements are resampled based on the conventional particle filter to reduce the influence of the particle impoverishment in every iterative estimation step due to enriching particle diversity. The simulated analysis of the proposed algorithm was conducted, while the algorithm was also utilized for the underwater small target detection experimental data processing. The results show that the mixture particle filter algorithm acquires state estimation with more stability and less error comparing to conventional particle filter. The mixture particle filter is effective to improve accuracy and stability for underwater small target tracking.