雷达多目标交叉轨迹跟踪算法

发布时间：2018-06-17 05:16

本文选题：粒子滤波 + 多目标跟踪　；参考：《南京大学学报(自然科学)》2017年04期

【摘要】：针对雷达跟踪系统中的多目标运动情况,以及目标轨迹交叉重叠导致的目标丢失或跟踪错误等问题,将多目标联合状态下的加权重采样思想引入到标准粒子滤波算法中.对基本粒子滤波算法进行优化,使用离散随机变量模拟目标后验概率,存在多个目标时引入联合状态概念,用关联函数把多个目标的状态变量和观测变量表示出来,把联合状态下的采样数据加入跟踪粒子的权值更新过程,使得粒子能够根据目标间的状态变化准确预测各个目标的后验分布,从而更新目标运动估计值,避免下一时刻粒子采样分布错误.联合加权重采样优化后的粒子能够准确跟踪目标运动,粒子预测和更新也不受目标交叉运动影响,克服跟踪不稳定或跟踪误差较大的问题.仿真结果表明,改进的粒子滤波算法能够达到正确跟踪多个目标的效果,并且目标轨迹交叉时仍然能够保持较高的跟踪精确度.
[Abstract]:In order to solve the problems of multi-target motion in radar tracking system and target loss or tracking error caused by cross-overlapping of target trajectory, the weighted resampling method in multi-target joint state is introduced into the standard particle filter algorithm. The basic particle filter algorithm is optimized, the posterior probability of the target is simulated by discrete random variables, the concept of joint state is introduced when there are multiple targets, and the state variables and observation variables of multiple targets are expressed by correlation function. The sampling data in the joint state are added to the updating process of the weight value of the tracking particle, so that the particle can accurately predict the posterior distribution of each target according to the state change between the targets, and then update the estimated value of the target motion. Avoid the error of particle sampling distribution at the next moment. The optimized particle with combined weighted sampling can track the target accurately, and the prediction and update of the particle is not affected by the cross-motion of the target, which can overcome the problem of unstable tracking or large tracking error. The simulation results show that the improved particle filter algorithm can track multiple targets correctly, and the tracking accuracy is still high when the trajectory of the target is crossed.
【作者单位】：南京大学电子科学与工程学院通信工程系;
【基金】：江苏省基础研究计划(BK20151391)
【分类号】：TN953

【相似文献】