自适应神经网络的光电跟踪成像消旋控制系统

为解决机载光电跟踪系统在跟踪过程中由于工作平台框架的转动导致成像画面的旋转问题,提出了一种基于自适应神经网络的消旋控制方法。系统以消旋指令角作为给定位置信息,以光电编码器实测角度值前后两拍之差作为实测速度值,组成速度反馈内环;以陀螺仪测得的角度值作为位置反馈值,构成位置外环;校正算法采用二阶超前-滞后校正并加入了自适应神经网络算法对其控制参数进行自适应调整。实验结果表明,在消旋拍摄过程中,消旋速度满足设计要求,拍摄图片清晰,消旋精度(均方值)达到1.4’,比传统校正方法输出误差减少了46%。

Despun control system for electro-optical tracking imaging based on adaptive neural network

In order to solve the problem of the rotation for imaging picture caused by the azimuth direction change of optical system and imaging device,a despun control system based on adaptive neural network is presented.The despun instruction angle is used as the given position in the system.The difference between two practical angle values obtained by photoelectric encoder is used as the measured velocity value,which is the velocity feedback inner loop.The angle value measured by gyroscope is used as the position feedback value,which is the position outer loop.A second order lead-lag correction method is used in this system,and an adaptive neural network algorithm,which adjusts the control parameters,is also used.The experimental results show that the despun velocity meets the design requirement in shooting.The picture shot is clear.And the despun accuracy(mean square) is 1.4′which decreases by 46% than the traditional error correction method.