摘要
研究无人机捷联导航姿态精度优化问题,针对微小型无人机做连续大机动飞行时,MEMS器件用于载体航姿测量精度低、易发散的问题,提出了一种基于UKF技术的姿态融合算法。用重力加速度在机体系的分量和陀螺漂移做为待估的状态量,建立了非线性的滤波模型。在系统模型噪声为复杂加性噪声且量测方程为线性方程时,推导出简化UKF算法。为了验证上述算法的有效性,将UKF和EKF算法进行对比,并通过姿态误差均值和均方差对实验结果进行定量分析。仿真结果表明,数据融合判别准则合理可行,改进算法提高了载体机动情况下的姿态精度,达到了预期的要求。
A new kind of attitude fusion algorithm based on UKF is presented when MEMS devices used in Vehi- cle navigation suffer the problems of low precision and divergent characteristics mostly when micro UAVs are doing continuous large maneuvering flight. A nonlinear filtering model was introduced with acceleration components of the earth gravity in the body frame and gyro drift as Kalman state vector. A simplified UKF（ unscented Kalman filter） fil- tering algorithm was derived under the situation where process noise is complex additive noise with linear measure- ment equation. In order to verify the effectiveness of the algorithm in this paper, the paper compared UKF and EKF and then gave quantitative analysis of the experiment＇s result through the mean of attitude error and mean square er- ror. The simulation results show that criterions of data fusion are reasonable and feasible, and the attitude accuracy of the algorithm can satisfy the application demand in mobile status of the carrier, which reaches the expected effect.
出处
《计算机仿真》
CSCD
北大核心
2014年第8期41-44,77,共5页
Computer Simulation
基金
航空基金(20125853035)
关键词
航姿系统
无迹卡尔曼滤波
非线性
判别准则
Attitude and heading reference system
Unscented kalman filter（UKF）
Nonlinear
Criterion
