应用双重窗口的储运发箱行为姿态检测方法
Dual-Window-Based Method for Detecting Behavioral Attitudes of Storage,Transportation and Launch Containers
张伟 1邓士杰 1于贵波1
作者信息
- 1. 陆军工程大学 石家庄校区,河北 石家庄 050003
- 折叠
摘要
储运发箱作为火箭弹的承载装备,在转运过程中,对其姿态进行监控和安全状态评估具有重要意义.为准确识别储运发箱在转运过程中的运动变化,设计了一种应用双重窗口的储运发箱行为姿态检测方法.通过姿态传感器获取储运发箱的三轴运动数据;使用第一重窗口对时序数据进行时间分割和特征提取,并结合时序卷积网络(temporal convolutional network,TCN)完成对储运发箱所处阶段的识别;使用第二重窗口对储运发箱可能会发生的危险行为进行判断;综合双重窗口的结果完成对储运发箱行为姿态的识别.实验结果表明,该方法能够准确地识别储运发箱在转运过程不同阶段的危险行为,为未来智能化储运发箱的研制提供了方法参考,具有一定的工程应用价值.
Abstract
The storage,transportation and launch container,serving as the carrying equipment for rocket projectiles,holds significant importance in monitoring its attitude and assessing its safety status during transit.To accurately identify the motion changes of the storage,transportation and launch con-tainer during transit,a dual-window-based method for detecting its behavioral attitudes is designed.Three-axis motion data of the container is obtained through attitude sensors;the first window is used for time segmentation and feature extraction of the time-series data,combined with the Temporal Convolu-tional Network(TCN)to identify the phase in which the container is located;the second window is uti-lized to assess potential hazardous behaviors of the container;and the results from both windows are inte-grated to complete the recognition of the container's behavioral attitudes.The experimental results demon-strate that this method can accurately identify hazardous behaviors of the storage,transportation and launch container at different stages during transit,providing a methodological reference for the future de-velopment of intelligent storage,transportation and launch containers and possessing certain engineering application value.
关键词
储运发箱/行为姿态/双重窗口/时序卷积网络Key words
storage,transportation and launch containers/behavioral attitudes/dual windows/Temporal Convolutional Network(TCN)引用本文复制引用
出版年
2024
NETL
NSTL
万方数据