MEMS INS/GNSS组合导航系统在车辆和机器人载体上的性能对比分析  

Comparative Analysis of Performance of MEMS INS/GNSS Integrated Navigation System on Civilian and Robotic Vehicle

作　　者：樊静 陈起金 唐海亮 牛小骥[1] FAN Jing;CHEN Qijin;TANG Hailiang;NIU Xiaoji(GNSS Research Center,Wuhan University,Wuhan Hubei 430079,China)

机构地区：[1]武汉大学卫星导航定位技术研究中心,湖北武汉430079

出　　处：《传感技术学报》2020年第8期1216-1222,共7页Chinese Journal of Sensors and Actuators

基　　金：国家重点研发计划项目(2016YFB0501800)。

摘　　要：同一款MEMS INS/GNSS组合导航系统在不同运动载体上的定位定姿性能差异会显著影响应用效果,有必要做深入的定量研究。本文首先采用奇异值分解方法,分析了组合导航状态量在载体不同运动场景下的可观测度。然后,对比了车辆和轮式机器人载体在开阔场景和GNSS卫星失锁场景下的组合导航性能,并详细分析了两种载体在不同运动模式下的导航误差和卡尔曼滤波均方误差。实测结果表明,车载和轮式机器人组合导航性能的差异主要表现在航向角方面;载体的加减速和转弯有助于航向角和陀螺z轴零偏误差的收敛;轮式机器人采用自动指令控制,造成运动模式单一,加减速时段少,同时转弯过程中加速度变化幅度小,转弯对其航向角和陀螺z轴零偏误差的收敛作用不如车载明显,因此,轮式机器人航向角误差比车载大,差异显著。An MEMS INS/GNSS integrated system may have significantly different navigation performance when mounted on different type host vehicles,which deserves a quantitative analysis. Firstly,a singular value decomposition(SVD)based method is used to analyze the observability of the navigation states under different vehicular dynamics. Secondly,the navigation performance on a civilian and a robotic vehicle are compared for open-sky cases and GNSS signal outage cases. Besides,the navigation error and Kalman filter mean square error of the two vehicles are analyzed in detail according to different motion modes. Results show that the difference in heading angle is most significant for these two host vehicles. Speed change and turning of the vehicle would benefit the heading convergence and the estimation of z-gyroscope’s bias. In addition,since the wheeled robotic vehicle generally experiences weaker maneuver and smaller acceleration or deceleration,the heading and z-gyro’s bias estimate convergence is not as obvious as that for an civilian host vehicle. As a result,the yaw error of wheeled robot is greater than that of vehicle.

关 键 词：INS/GNSS组合导航 MEMS惯导 轮式机器人 姿态误差 可观测度分析 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]