基于空间移动平台的空间激光通信是实现未来超大容量空间通信的主要途径, 而高精度空间平台动态跟踪问题一直是其研究的难点之一。空间移动平台跟踪是一个纯角度的非线性跟踪, 用单一的模型很难准确描述空间移动平台的运动状态, 需要采用多个模型描述机载平台的运动状态, 但交互多模型估计(IMM)需用先验知识和复杂计算量, 研究将模型转移概率与模糊推理相结合解决测量空间的不确定性和复杂计算量。标准粒子滤波(PF)能够处理非线性问题, 但会引起粒子滤波器粒子退化和发散, 采用无迹粒子滤波(UPF)克服了标准粒子滤波没有考虑最新量测信息的问题。仿真结果表明模糊交互多模型粒子滤波算法能较大地提高跟踪性能。

Space laser communication link based on space mobile platform is the main method for achieving large capacity space communications, but high precision dynamic tracking of space mobile platform is always one of the difficult problems to be resolved. It is impossible to accurately describe the movement of airborne platform with the simple model, so it is necessary to adopt interacting multiple models (IMM), while traditional IMM need to know the priori knowledge and complex calculation, so in this paper, fuzzy IMM based on fuzzy logic inference can be used to replace model probability and decrease calculation complexity. Standard particle filter (PF) can deal with nonlinear/non-Gaussian problems, but it can cause the degeneracy phenomenon and sample impoverishment. Improved PF-unscented particle filter(UPF) can combine with current observations, which moves sample particles to the maximal posterior density estimation of the target state. Simulation results demonstrate the improved performance of the proposed method.