论文部分内容阅读
提出了一种改进的自适应新生目标强度的概率假设密度(PHD)滤波算法.首先,对归一化因子进行了分析,在此基础上,提出了一种改进滤波策略,有效解决了归一化失衡问题;其次,在量测点附近通过无迹变换(UT)产生样本点,然后再采用粒子群(PSO)算法寻找最优点,从而得到新生目标概率密度函数的近似估计;最后,在序列蒙特卡罗(SMC)框架下对算法进行了实现.采用一种回溯策略,通过记录新生目标的状态和数目,修正存活目标的估计数目和相关航迹,进而得到每个目标的完整航迹.仿真结果表明:改进算法可以有效跟踪多个机动目标的状态和数目,滤波精度较高,具有较好的工程应用前景.
This paper proposes an improved probability hypothesized density (PHD) filtering algorithm for adaptive nascent target intensity.Firstly, the normalization factor is analyzed, and on this basis, an improved filtering strategy is proposed to solve the problem of normalization Secondly, the sample points are generated through no-trace transform (UT) near the measuring point, and then the particle swarm optimization (PSO) algorithm is used to find the optimal point to obtain the approximate estimate of the probability density function of the nascent target. Finally, The algorithm is implemented in the framework of Monte Carlo (SMC), which uses a backtracking strategy to correct the number of survivors’ targets and their associated trajectories by recording the states and numbers of new targets, so as to obtain the complete trajectory of each target. The simulation results show that the improved algorithm can effectively track the state and number of multiple maneuvering targets with high filtering accuracy and good engineering application prospect.