论文部分内容阅读
针对粒子滤波的粒子退化和匮乏等问题,在粒子滤波和网格近似的基础上,提出了一种采用分层近似策略的粒子滤波改进算法。改进算法利用高斯分布对后验概率密度进行近似,并在连续分布的后验概率密度上进行分层近似,从而获得更具效率的粒子,提高了粒子滤波的精度。光纤陀螺的寻北精度主要取决于光纤陀螺自身的性能以及所采用的寻北方案,将采用分层近似策略的改进粒子滤波算法应用于光纤陀螺的寻北方案中,能够有效地解决光纤陀螺寻北中的非线性状态估计问题,提高光纤陀螺的寻北精度。
Aiming at the problem of particle filtering such as particle degradation and starvation, a particle filter based on particle filter and mesh approximation is proposed. The improved algorithm uses the Gaussian distribution to approximate the posterior probability density, and stratifies the posterior probability density of the continuous distribution to obtain more efficient particles and improve the accuracy of the particle filter. The north-seeking accuracy of optical fiber gyroscope mainly depends on the performance of optical fiber gyroscope and the north-seeking scheme adopted, and the improved particle filter algorithm using stratification approximation method is applied to the north-seeking scheme of optical fiber gyroscope, which can effectively solve the problem of fiber optic gyroscope North of the nonlinear state estimation problem, improve the search accuracy of fiber optic gyroscope.