金仁成 魏巍 孙庆飞 洪金辉 姜佳旭

摘要：针对传统磁力计/MIMU组合导航系统存在的抗电磁干扰能力弱、适应性差等问题，提出了一种基于大气偏振分布模式的无人机导航航姿无约束优化求解方法。融合陀螺仪、加速度计、偏振光传感器测量数据，构建基于改进拟牛顿法的姿态最优化解算模型，其中，该模型以最优化理论为基础，引入实时四元数微分构建动态步长因子，预测姿态更新趋势;设计状态判断阈值，自适应判断飞行状态，提高解算精度。实验结果表明，这种方法得到的姿态角静态性能稳定，同时，在动态环境中，该方法得到的姿态角误差值最小，较梯度下降法精度提高50%，较共轭梯度法精度提高37%～50%，表现出较好的动态收敛性与误差稳定性，动态精度较高，能够较准确跟踪姿态变化，可以有效提高偏振光辅助的捷联惯导组合导航系统航姿估计的准确性与可靠性。

关键词：偏振光;MIMU;航姿解算;拟牛顿

中图分类号：V249             文献标志码：A

文章编号：1009-9492（2021）12-0023-06

开放科学（资源服务）标识码（OSID）：

MIMU/ Polarized Light Quasi-newtonian Optimization Method for Attitude Solving

Jin Rencheng ，Wei Wei ，Sun Qingfei ，Hong Jinhui ，Jiang Jiaxu

（School of Mechanical Engineering， Dalian University of Technology， Dalian， Liaoning 116024， China）

Abstract： Aiming at the problems of the traditional magnetometer /MIMU integrated navigation system， such as weak anti-electromagnetic interference  ability  and  poor  adaptability，  an  unconstrained  optimization  method  for  UAV  navigation  attitude  was  proposed  based  on atmospheric polarization distribution mode. The attitude optimization model based on the improved quasi-Newton method was constructed by integrating the measurement data of gyroscope， accelerometer and polarized light sensor. Based on the optimization theory， the dynamic step factor was constructed by introducing real-time quaternion differential to predict the attitude updating trend. The threshold value of state judgment was designed， and the flight state was judged adaptively to improve the precision of solution. Attitude angle of the experimental results show that the approach of static performance is stable， at the same time， in a dynamic environment， the method to get the minimum value attitude angle error， the gradient descent method precision is increased by 50%， the conjugate gradient method precision is increased by 37%～50%， show a good dynamic stability， convergence and error dynamic accuracy is higher， can accurately track the attitude change， it can effectively improve the accuracy and reliability of attitude estimation of strapdown inertial navigation system assisted by polarized light.     Key words： polarized light; MIMU; navigation position calculating; quasi-newton

0 引言

隨着微机电系统（MEMS）工艺和计算机技术的发展，小型无人机多采用低成本的 MEMS传感器（主要有陀螺仪、加速度计、磁力计）作为姿态估计设备。其中，陀螺仪是姿态角度估计的关键传感器，但由于其固有误差，长时间累计积分运算，会存在严重漂移现象，通常利用加速度计补偿陀螺仪的俯仰角和横滚角偏差，利用磁强计补偿航向角偏差。但是磁强计易受地磁场与电磁场的干扰，不能用于某些特殊环境。……