【摘要】 针对高动态环境下惯性/天文组合导航精度下降的问题,提出一种基于神经网络辅助的惯性/天文组合导航方法。首先以组合导航滤波估计过程中的增益矩阵和动态环境下的惯性器件量测信息构建特征向量;然后,采用导航估计误差对BP神经网络进行训练;最后,利用BP神经网络的输出结果辅助修正组合导航系统。计算机仿真验证结果表明,相较于传统方法,基于BP神经网络辅助的惯性/天文组合导航系统的姿态估计精度可提高30%以上,在动态环境下姿态精度可以保持在5″(1σ)以内。所提出的方法对提高动态环境下惯性/天文组合导航系统的精度和适应能力具有一定的参考价值。更多还原

【Abstract】 Aiming at the problem that the precision of inertial/celestial navigation(ICN) is decreased in high dynamic environment, an ICN method based on neural network is proposed. First, the eigenvectors of BP neural network are constructed by the gain matrix in the ICN filtering estimation process and the angular velocity in dynamic environment. Then, the BP neural network is trained by navigation estimation errors. Finally, the filtering results of the ICN system is corrected by the output of BP neural network. Simulation results show that, by the proposed method, the attitude estimation accuracy of the ICN system can be improved by more than 30% compared with the traditional method, and the attitude accuracy can be maintained within 5?(1?) in high dynamic environment, which has important reference value for improving the accuracy and adaptability of the ICN in dynamic environment.更多还原