《计算机应用》唯一官方网站 ›› 2024, Vol. 44 ›› Issue (5): 1401-1407.DOI: 10.11772/j.issn.1001-9081.2023121837

所属专题: 进化计算专题(2024年第5期“进化计算专题”导读,全文即将上线)

• 进化计算专题 • 上一篇    

基于有限忍耐度鸽群优化的无人机近距空战机动决策

郑志强, 段海滨()   

  1. 北京航空航天大学 自动化科学与电气工程学院,北京 100083
  • 收稿日期:2024-01-02 接受日期:2024-01-22 发布日期:2024-04-26 出版日期:2024-05-10
  • 通讯作者: 段海滨
  • 作者简介:郑志强(1998—),男,江西九江人,博士研究生,主要研究方向:无人机自主控制、群体智能
    第一联系人:段海滨(1976—),男,山东广饶人,教授,博士,主要研究方向:无人机自主控制、计算机仿生视觉、智能感知、仿生智能计算。
  • 基金资助:
    国家自然科学基金资助项目(T2121003);科技创新2030—“新一代人工智能”重大项目(2018AAA0102403)

Short-range UAV air combat maneuver decision-making via finite tolerance pigeon-inspired optimization

Zhiqiang ZHENG, Haibin DUAN()   

  1. School of Automation Science and Electrical Engineering,Beihang University,Beijing 100083,China
  • Received:2024-01-02 Accepted:2024-01-22 Online:2024-04-26 Published:2024-05-10
  • Contact: Haibin DUAN
  • About author:ZHENG Zhiqiang, born in 1998, Ph. D. candidate. His research interests include UAV autonomous control, swarm intelligence.
  • Supported by:
    National Natural Science Foundation of China(T2121003);Science and Technology Innovation 2030 — Key Project of “New Generation Artificial Intelligence”(2018AAA0102403)

摘要:

由于对抗双方态势的快速变化,无人机近距空战机动自主决策困难且复杂,是空中对抗的一个难点。对此,提出一种基于有限忍耐度鸽群优化(FTPIO)算法的无人机近距空战机动决策方法。该方法主要包括基于机动动作库的对手行动预测和基于FTPIO算法的机动控制量和执行时间优化求解两个部分。为提升基本鸽群优化(PIO)算法的全局探索能力,引入有限忍耐度策略,在鸽子个体几次迭代中没有找到更优解时对其属性进行一次重置,避免陷入局部最优陷阱。该方法采用的优化变量是无人机运动模型控制变量的增量,打破了机动库的限制。通过和极小极大方法、基本PIO算法和粒子群优化(PSO)算法的仿真对抗测试结果表明,所提出的机动决策方法能够在近距空战中有效击败对手,产生更为灵活的欺骗性机动行为。

关键词: 鸽群优化算法, 近距空战, 机动决策, 无人机, 有限忍耐度策略

Abstract:

Due to the rapid change of the situation during the confrontation, the autonomous maneuver decision-making for short-range Unmanned Aerial Vehicle (UAV) air combat is difficult and complex, which is a difficult point in air combat. To address this issue, a short-range UAV air combat maneuver decision-making method based on Finite Tolerance Pigeon-Inspired Optimization (FTPIO) algorithm was proposed. Two parts were designed in the proposed method: opponent action prediction based on the maneuver library and optimization solution of maneuver control and execution time based on FTPIO algorithm. To improve the global exploration ability of the basic Pigeon-Inspired Optimization (PIO) algorithm, the finite tolerance strategy was introduced. When the pigeon individual failed to find a better solution in several iterations, its attributes would been reset to avoid falling into the local optimal trap. The optimization variables used in the proposed method were the increments of the control variables of UAV motion model, which broke the limitations of the maneuver library. Simulation and adversarial testing results with the MiniMax method, basic PIO algorithm, and Particle Swarm Optimization (PSO) algorithm show that the proposed maneuver decision-making method can effectively defeat opponents during confrontation and generate more flexible deceptive maneuver behaviors.

Key words: Pigeon-Inspired Optimization (PIO) algorithm, short-range air combat, maneuver decision-making, Unmanned Aerial Vehicle (UAV), finite tolerance strategy

中图分类号: