现代防御技术 ›› 2024, Vol. 52 ›› Issue (1): 49-56.DOI: 10.3969/j.issn.1009-086x.2024.01.007

• 军事智能 • 上一篇    下一篇

基于改进支持向量回归的空战飞行动作识别

刘庆利(), 李蕊, 乔晨昊   

  1. 大连大学 通信与网络重点实验室,辽宁 大连 116622
  • 收稿日期:2022-12-07 修回日期:2023-03-03 出版日期:2024-02-28 发布日期:2024-02-21
  • 通讯作者: 李蕊
  • 作者简介:刘庆利(1981-),男,辽宁大连人。教授,博士,研究方向为计算机网络、智能指控系统、智能决策系统。E-mail:liuqingli@dlu.edu.cn

Air Combat Flight Action Recognition Based on Improved Support Vector Regression

Qingli LIU(), Rui LI, Chenhao QIAO   

  1. Key Laboratory of Communication and Network,Dalian University,Dalian 116622,China
  • Received:2022-12-07 Revised:2023-03-03 Online:2024-02-28 Published:2024-02-21
  • Contact: Rui LI

摘要:

针对空战中飞机的飞行动作愈发复杂导致识别准确率低的问题,提出了改进支持向量回归的空战飞行动作识别方法,该方法采用高斯核函数作为线性核函数,利用混沌初始化和反向学习策略改进麻雀搜索算法,利用改进后的麻雀算法优化支持向量回归算法,具体表现为对支持向量回归算法中高斯核函数的参数进行优化,通过优化后的支持向量回归算法进行飞机动作识别。采用了五种基本的飞行动作和几种复杂的飞行动作验证该方法的识别准确率。仿真表明,优化后的支持向量回归算法与传统的支持向量回归算法、模糊支持向量机算法、传统聚类算法、神经网络算法相比,对基本飞行动作的平均识别率至少提升了2.2%,对复杂飞行动作的平均识别率至少提升了3.7%。

关键词: 空战, 支持向量回归, 强化麻雀搜索算法, 飞行动作识别, 复杂动作

Abstract:

Aiming at the problem of low recognition accuracy due to the increasing complexity of aircraft flight movements in air combat, this paper proposes an air combat flight situation recognition method based on enhanced support vector regression. The sparrow search algorithm is improved by using pinhole imaging and chaos initialization. The improved sparrow algorithm is used to optimize the support vector regression algorithm, which is specifically represented by the optimisation of the parameters of the Gaussian kernel function in the support vector regression algorithm. The optimized support vector regression algorithm is used to identify aircraft movements.Five basic flight actions and complex flight actions are used to verify the recognition accuracy of the method. Simulation shows that the optimised support vector regression algorithm improves the average recognition rate of basic flight manoeuvres by at least 2.2%, and the average recognition rate of complex flight action by at least 3.7%, compared with the traditional support vector regression algorithm, fuzzy support vector machine algorithm, traditional clustering algorithm, and neural network algorithm.

Key words: air combat, support vector regression, intensify sparrow search algorithm, flight action recognition, complex action

中图分类号: