稀疏网格求积分滤波算法在SINS/GPS紧组合导航中的应用（英文）

Application of sparse grid quadrature filter to tightly-coupled SINS/GPS integrated navigation system

高精度的导航信息对于高空飞行器至关重要。针对高空飞行器的特点,选取发射点惯性坐标系为导航坐标系,建立基于伪距、伪距率的SINS/GPS紧组合导航系统数学模型。针对该系统的状态方程和量测方程非线性的特性,采用基于稀疏网格求积分滤波算法。整个设计实现了对准与导航的一体化,避免了将对准与导航分别设计的繁琐过程。仿真结果表明,在飞行器起飞阶段,由于系统的非线性较强,稀疏网格求积分滤波算法比UKF滤波算法的对准精度更高,并且对准速度更快;通过比较稀疏网格求积分滤波算法在不同组合方式下的估计效果,可以看出采用紧组合方式可以明显提高导航精度。最后采用不同精度的传感器进行仿真,结果表明基于稀疏网格求积分滤波算法的紧组合算法能够适用的传感器精度范围较广。

High-precision navigation information is crucial for high altitude vehicles. Considering the characteristics of high-altitude aircrafts, we select the launch inertial coordinate system as the navigation coordinate system and propose a mathematic model for tightly-coupled SINS/GPS integrated navigation system based on pseudo-range and pseudo-range rate. As the state equations and measurement equations are nonlinear, the sparse grid quadrature filter (SGQF) is adopted. The method proposed in this paper is fit for both aligning and navigating, so it is more efficient compared with the method that designs aligning and navigating separately. Simulation results indicate that, owing to this strong nonlinear system, the sparse grid quadrature filter can not only estimate navigation parameters faster but also more accurately than the unscented Kalman filter (UKF) during the take-off phase of high-altitude aircrafts. They also show that the sparse grid quadrature filter with tightly-coupled integration can greatly improve estimation accuracy of navigation compared with that with loose integration algorithm. Finally, the impact of different levels of accuracy of inertial devices is studied. The result indicates that tightly-coupled integration with SGQF can perform quite well within a large range of accuracy of inertial devices. ©, 2014, Editorial Department of Journal of Chinese Inertial Technology. All right reserved.