论文部分内容阅读
针对基本粒子滤波(PF)算法存在的粒子退化和重采样引起的粒子多样性丧失,导致粒子样本无法精确表示状态概率密度函数真实分布,提出了一种基于混沌的改进粒子群优化(PSO)粒子滤波算法。通过引入混沌序列产生一组混沌变量,将产生的变量映射到优化变量的区间提高粒子质量,并利用混沌扰动克服粒子群优化局部最优问题。利用单变量非静态增长模型(UNGM)在高斯噪声和非高斯噪声环境下将该算法与基本粒子滤波和粒子群优化粒子滤波(PSO-PF)的性能进行仿真比较。结果表明:该算法的性能在有效粒子数和均方根误差(RMSE)等参数都优于基本粒子滤波和粒子群优化粒子滤波,改善了算法的精度和跟踪性能。
Aiming at the loss of particle diversity caused by particle degeneration and resampling existing in the basic particle filter (PF) algorithm, the particle sample can not accurately represent the true distribution of the state probability density function. A modified Particle Swarm Optimization (PSO) particle based on chaos is proposed Filtering algorithm. By introducing a chaotic sequence, a set of chaotic variables is generated. The generated variables are mapped to the optimal variables to improve the particle quality. The chaotic perturbation is used to overcome the particle swarm optimization problem. The performance of this algorithm is compared with that of particle swarm optimization and particle swarm optimization Particle Swarm Optimization (PSO-PF) using Gaussian noise and non-Gaussian noise by univariate non-stationary growth model (UNGM). The results show that the performance of the proposed algorithm is better than the basic particle filter and the Particle Swarm Optimization (PSO) particle filter in terms of effective particle number and root mean square error (RMSE), which improves the accuracy and tracking performance of the algorithm.