融入粒子群优化的UPF算法研究及其导航应用

Research on a Novel UPF Algorithm Incorporatingthe PSO and the Application on TAN

粒子滤波易出现粒子多样性损失,粒子退化等问题,且在初始未知时,需要粒子数较多,收敛较慢。针对上述缺陷,在UKF和PF滤波基础上,将改进了搜索因子的粒子群优化算法融入U粒子滤波算法,提出了一种新型粒子滤波算法,加速了算法的收敛速度。将其应用于地形匹配导航算法中,仿真结果表明该新型算法明显优于UPF及PF算法,更适应于大误差的初始条件下,收敛速度快,具有更高的导航精度和抗噪特性。

The disadvantages of the particle filter are the degeneracy of particles,and the difficulty in sampling from the posterior probability distributing.In addition,it needs a large number of sample particles while the initial state is unknown.To solve the problems referred above and based on the UKF methods,a novel optimized PF algorithm is represented.The novel algorithm incorporates the UPF with the Particle swarm optimized method which advances the searching factor,hence the particles are moved towards regions where they have larger value according to the posterior density function.At the end,the simulation of the novel algorithms applying on the AUV terrain navigation is done,the result indicates that the novel method has higher accuracy of the navigation and better avoidance of the noise.