基于改进Agent的空域管制扇区划设方法  

Improved Agent-Based Sectorization Method of Air Traffic Management

作　　者：支博 田勇[1] 梁满佳 黄潇 吕越 ZHI Bo;TIAN Yong;LIANG Man-jia;HUANG Xiao;LV Yue(Civil Avaition College,Nanjing University of Aeronautics and Astronautics,Nanjing 211106,China)

机构地区：[1]南京航空航天大学民航学院,南京211106

出　　处：《科学技术与工程》2025年第7期3035-3043,共9页Science Technology and Engineering

基　　金：江苏省青年科学基金(BK20230892);江苏省双创博士人才项目(JSSCBS20220212)。

摘　　要：为解决传统基于主观经验人为划设的扇区所存在的空域资源分配不均问题,满足如今空中交通运行需求,通过改进Agent方法,研究终端区三维扇区划设问题。首先,在满足传统扇区划设约束条件下,以提高扇区同交通流适应性,降低并均衡管制负荷为目标;随后,通过使用遗传算法确定Agent初始解位置以提高计算效率、设计优化Agent生长规则与空域填充规则方式改进传统Agent方法;最后,以上海终端区内数据进行实例分析验证,结果表明:相比现行与传统Agent方法所得扇区划设方案,改进Agent方法能够分别提升25.84%、18.54%的扇区形状特征与空中交通流适应性;同时在降低终端区管制负荷总和基础上,可将各扇区间管制负荷标准差分别降低53.33%、36.58%。可见基于改进Agent的空域管制扇区划设方法为中国本土特色化空域规划提供参考性意见。In order to solve the problem of uneven allocation of airspace resources in traditional artificial sectors based on subjective experience,and to meet the needs of today s air traffic operation,the problem of three-dimensional sectorization in terminal areas was studied by improving Agent method.Firstly,while adhering to traditional sectoring constraints,the objective was to enhance sector adaptability to traffic flows and achieve a reduction and balance in air traffic control workload.Subsequently,the traditional Agent method was improved by using genetic algorithm to determine the location of Agent initial solution,so that it could enhance computational efficiency,designing and optimizing Agent growth rules and spatial filling rules.Finally,using the Shanghai terminal area as a case study,the results indicated that the improved Agent method yields sector planning scheme with respective improvements of 25.84%and 18.54%in sector shape characteristics and adaptability to airborne traffic flows.Simultaneously,while reducing the overall terminal area air traffic control workload,the standard deviation of control workload among sectors was reduced by 53.33%and 36.58%,respectively,compared to the existing and traditional Agent methods.It can be seen that the Research on Improved Agent-Based Sectorization Method provides reference for the local characteristic airspace planning of our country.

关 键 词：空中交通管理 扇区划设 AGENT模型 启发式算法 强化学习 

分 类 号：V355[航空宇航科学与技术—人机与环境工程]