基于随机森林的HRGV滑翔段飞行状态识别

针对临近空间高超声速再入滑翔飞行器(Hypersonic Reentry-Glide Vehicle,HRGV)滑翔段飞行状态识别问题,提出了一种基于随机森林的识别方法。首先将目标的飞行状态分为6类,通过运动方程生成具有代表性的飞行数据;其次分析了目标运动特性,利用运动参数构造特征属性并对其进行处理和筛选,得到最终样本数据。为实现雷达跟踪轨迹的飞行状态识别,将雷达跟踪数据进行平滑处理,经坐标变换后得到运动参数估计结果,并用训练好的分类器识别目标的飞行状态。试验结果表明,所设计的随机森林分类器识别精度较高,但当运动参数的估计存在误差时,识别精度会有一定程度的下降。

Flight State Recognition of HRGV Glide Section Based on Random Forest

Aiming at the problem of flight state recognition of hypersonic reentry-glide vehicle( HRGV)glide section in near space,a recognition method based on random forest is proposed. Firstly,the flight states of the target are divided into six categories,and representative flight datas are generated by motion equation. Secondly,the motion characteristics of the target are analyzed. Then,the feature attributes are constructed by motion parameters,and processed and filtered to obtain the final sample data. In order to realize the flight state recognition of radar tracking trajectory,the radar tracking data is smoothed,and the motion parameters are estimated by coordinate transformation. The trained classifier is used to identify the flight state of the target. The experimental results show that the recognition accuracy of the designed random forest classifier is high. But when the estimation of motion parameters has errors,the recognition accuracy will decrease to a certain extent.