摘要
目标运动估计是光电跟踪平台实现共轴控制的核心。针对目标运动的滤波精度问题,提出了基于灰色预测模型和卡尔曼理论的目标运动融合滤波方法。利用少量前几个时刻的目标运动数据实时在线建立灰色预测模型,替代标准卡尔曼滤波的目标状态预测方程,避免了需要预先假定目标运动数学模型而引起滤波误差的弊端。进一步对灰色模型预测值和融合滤波估计值的残差序列,构建残差在线预估模型;利用残差预测值对灰色预测模型进行实时修正,有效地提高了灰色模型的预测精度,改善了融合滤波的效果。仿真实验验证了灰色模型在目标运动预测中的有效性。实验结果表明:残差修正策略将灰色模型预测精度提高了60%;与其他滤波方法相比,提出的融合滤波方法具有更好的滤波效果及更准确的目标运动估计。
Target motion estimation is the key factor of on-axis control of electro-optical-tracking platform.For the filtering accuracy of target motion,a fusion filtering method based on grey prediction model and Kalman theory is proposed.A grey prediction model is set up online in real time by using a small amount of target motion in the first few moments and replaces the target state prediction equation of Kalman filter,which avoids some filtering error due to presupposion of target motion model.A residual prediction model is also built online by using the residual sequence which is generated by the predicted and estimated data.The predicted residual value is used to modify the grey prediction value in real time.The simulated results show that the prediction accuracy of the grey model based on residual modification can be increased by 60%;compared with some other filtering methods,the proposed fusion filtering method has better filtering effect and more accurate estimation.
作者
吕明明
刘荣忠
侯远龙
高强
王力
Lü Mingming;LIU Rongzhong;HOU Yuanlong;GAO Qiang;WANG Li(School of Mechanical Engineering,Nanjing University of Science and Technology,Nanjing 210094,Jiangsu,China)
出处
《兵工学报》
EI
CAS
CSCD
北大核心
2019年第3期548-554,共7页
Acta Armamentarii
基金
国家自然科学基金项目(51305205)
关键词
光电跟踪平台
共轴控制
卡尔曼滤波
灰色预测模型
残差修正
electro-optical tracking platform
on-axis control
Kalman filter
grey prediction model
residual modification
