弹道修正引信滚转角未知条件下MIMU快速初始对准方法

Rapid Initial Alignment of MIMU with Unknown Roll Angle of the Trajectory Correction Fuze

针对基于双旋结构的二维弹道修正引信及所配用弹药发射后修正引信滚转角不确定的问题,提出采用BP神经网络快速确定修正引信滚转角范围进行粗对准,进而利用卫星导航信息辅助惯性导航系统进行快速初始对准的方法.该方法利用修正引信滚转角误差与惯性导航解算速度误差之间的关系建立神经网络模型,并通过仿真生成不同工况的弹道数据对神经网络进行训练.针对某旋转弹弹道进行仿真,结果表明,使用中等精度微惯性测量组合,粗对准过程中修正引信滚转角误差估计值小于20°,精对准过程中姿态误差角在弹丸发射后40 s内收敛到0.4°以内.

To solve the problem of unknown roll angle of correction fuze in the two-dimensional trajectory correction projectile with dual-spin structure after launch, a method was proposed, using back propagation(BP)neural network to determine the roll angle range of correction fuze for coarse alignment, and then using the satellite navigation system to assist inertial navigation system for rapid initial alignment. Considering the relation between the roll angle error of correction fuze and velocity error of inertial navigation,a neural network model was established and trained with different simulated trajectories of multiple working conditions. The simulation results of spinning projectile trajectory show that the error of estimated roll angle of correction fuze is less than20° during coarse alignment process when using medium precision MIMU and the error can converge to less than 0.4° within 40 s after launch during fine alignment process.