量子PSO粒子滤波在DR/GPS组合导航系统中的应用

将量子粒子群优化(QPSO)算法与粒子滤波(PF)相结合,提出了量子PSO粒子滤波(QPSO-PF)算法,对航位推算(DR)与GPS组合导航系统中的里程系数误差和航向误差进行辨识估计,并对里程系数和航向进行修正。该算法采用量子位对粒子进行编码,引入量子旋转门与变异操作保持了粒子集的多样性,通过QPSO搜索寻优重新分配粒子,使粒子集有效地逼近真实的后验概率分布,从而有效地减轻了退化现象,提高了PF的精度。DR/GPS组合导航系统跑车实验结果表明:该算法有效地抑制了DR导航系统误差的增长,提高了组合导航系统的定位精度。

Application of QPSO-PF in DR/GPS integrated navigation system

Quantum PSO(QPSO)particle filter(QPSO-PF) is proposed by combing the QPSO algorithm with PF,by which the mileage coefficient error and azimuth error is estimated in DR/GPS integrated navigation system,and then the mileage coefficient and azimuth is corrected.In QPSO-PF,the quantum bit is used to encode the particle and the quantum rotating gate and mutation are introduced to guarantee the diversity of particle set,and the particles are redistributed by the optimization search of QPSO,and makes the particle set effectively approximate the true posterior probability distribution,which effectively alleviates the degenerative phenomenon and improves the precision of PF.The experimental results of vehicle experiment of DR/GPS integrated navigation system show the increase trend of error of DR navigation system is effectively restrained,and the positioning precision of integrated navigation system is effectively improved by using QPSO-PF.