摘要
针对探空火箭在高动态与非线性环境下高精度导航问题,基于DSP6747+FPGA双核构架,设计了一种低成本的新型GNSS/MEMS一体化紧组合导航系统;同时,在直接法滤波的基础上采用了改进的自适应渐消扩展卡尔曼粒子滤波(PF-AFEKF)算法,解决了粒子退化问题并充分利用了当前观测信息以及自适应权重因子抑制了滤波发散,实现了高动态、非线性状态下的GNSS/MEMS深度信息融合;最后,进行了地面跑车试验,试验结果证明了所设计的低成本GNSS/MEMS紧组合导航系统具有较高的导航精度,容错能力强。
Aiming at the high precision navigation problem of sounding rocket under high dynamic and nonlinear environment, a lowcost GNSS/MEMS tightly coupled integrated navigation system based on DSP6747 and FPGA dual core hardware platform is designed. In addition, an improved Particle Filter-Adaptive Fading Extended Kalman Filter （PF-AFEKF） algorithm based on direct mode is designed to achieve high performance of information fusion under high dynamic and nonlinear condition, and problems of particle degeneracy and filter divergence can be solved and restrained by taking ful! advantage of current observational information and the adaptive weight factor. Finally, vehicle experiment is conducted and experiment results show that the designed low-cost GNSS/MEMS navigation system has higher navigation precision and fault tolerance.
出处
《计算机测量与控制》
2016年第11期236-239,248,共5页
Computer Measurement &Control
基金
江苏省自然科学基金(BK20140795)
中国航空科学基金(20145159002)
中国航天CAST创新基金(CAST2014-27)
关键词
组合导航系统
紧组合
双核构架
非线性
integrated navigation system
tightly-coupled
dual core
nonlinear
