摘要
为改进管制工作负荷预测方法在探究主、客观量映射关系方面的局限性,提出了一种空中交通管制复杂度预测模型。定义空中交通管制复杂度为管制员工作负荷与其阈值之比;组织一线管制员及资深专家进行多组模拟机实验,获取其对各实验场景下管制复杂度的主观定性评估,利用MATLAB处理得到相应场景下的复杂因子;通过BP(back propagation)神经网络对样本数据进行非线性拟合。拟合模型的平均绝对误差为0.025,预测偏离程度为3.75%。研究表明该模型能够较准确地反映空中交通管制复杂度(主观量)与各复杂因子(客观量)之间的映射关系,为空域规划与管理提供科学理论支持。
In order to improve the limitation of control workload predictive method in exploring the mapping relationship between subjective and objective quantities,an air traffic control complexity prediction model is proposed.Firstly,the complexity of air traffic control was defined as the ratio of controller workload to its threshold.Secondly,front-line controllers and senior experts were organized to conduct multiple groups of simulator experiments to obtain their subjective and qualitative evaluation of control complexity in each experimental scenario.Meanwhile,the complexity factors in corresponding scenarios were obtained through MATLAB.Finally,back propagation(BP)neural network was used to fit the sample data.The average absolute error of the fitting model is 0.025,and the prediction deviation is 3.75%.The results show that this model can accurately reflect the mapping relationship between the complexity of air traffic control(subjective quantity)and various complexity factors(objective quantity),and provide scientific theoretical support for airspace planning and management.
作者
朱承元
惠雅婷
王毅鹏
ZHU Cheng-yuan;HUI Ya-ting;WANG Yi-peng(College of Traffic Management, Civil Aviation University of China, Tianjin 300300, China)
出处
《科学技术与工程》
北大核心
2021年第18期7790-7796,共7页
Science Technology and Engineering
基金
国家自然科学基金青年科学基金(U1833103)。
关键词
空中交通管制
空中交通管制复杂度
BP神经网络
雷达模拟机
air traffic control
air traffic control complexity
back propagation(BP)neural network
radar simulator
