针对粒子滤波存在粒子质量低与粒子退化的问题,提出了一种基于二阶自适应权值粒子滤波算法。将算法分为两个阶段,首先,多传感器数据发送给相应的粒子滤波计算模块,以优化粒子分布为目的更新建议分布密度;之后,在最终的自适应权值粒子滤波模块中对多传感器数据构造完整的似然函数,同时通过欧氏距离和反映量测噪声统计特性的精度因子进行自适应权值分布调整,最终得到更精确的估计。进行实例仿真分析,所得结果验证了该算法的有效性。
In order to solve the problems ol low quality and degeneration ol particles in the process ol particle filte -ring, a particle filtering algorithm based on the two-stage adaptive weight was proposed. The algorithm was divided into two steps. First, multi-sensor data was sent to the appropriate particle filter calculation module,with optimizing particle distribution as a purpose,and the proposed distribution density was updated. Second,in the final particle filter module based on the adaptive weight,a complete likelihood function was structured. Meanwhile,by Euclidean distance and the accuracy factor reflecting the statistic properties of the measured noise,the distribution of the adaptive weight was adjusted to attain a more accurate estimate. A simulation experiment shows the effectiveness of the algorithm.