摘要
针对执行较长飞行任务的飞行器在飞行任务期间难以实时准确预测机动能力的问题,开展了基于长短期记忆(LSTM)的飞行器纵向可用过载预测方法研究。首先,对飞行器纵向过载相关参量进行了分析。然后,以纵向可用过载为性能指标,建立了基于LSTM网络的BP神经网络预测模型。预测模型的输入是一段飞行时间内可测量的飞行状态数据序列,输出是未来时刻的纵向可用过载。最后,基于某型飞行器建立数字仿真模型并开展了仿真验证及结果分析。研究结果表明,所提出的预测模型准确有效,可以帮助实现飞行器飞行性能的实时评估和预测。
A research had been carried out on the prediction method of longitudinal available over-load for aircraft based on the long short-term memory(LSTM)network,aiming to address the issue of real-time and accurate prediction of maneuverability during the extended flight missions for air-craft.Firstly,an analysis of the parameters related to longitudinal overload of the aircraft was con-ducted.Then,taking the longitudinal available overload as the performance index,a BP neural net-work prediction model based on LSTM network was established.The input of the prediction model is the measurable flight status data series over a period of flight time,and the output is the longitudinal available overload at a future time.Finally,a digital simulation model was established based on a certain type of aircraft,and simulation verification and result analysis were carried out.The research results show that the proposed prediction model is accurate and effective,and it can help to achieve the real-time evaluation and prediction of the flight performance of aircraft.
作者
张琬琪
程月华
余自权
曹瑞
ZHANG Wanqi;CHENG Yuehua;YU Ziquan;CAO Rui(College of Automation Engineering,NUAA,Nanjing 211100,China;College of Information Engineering,Yangzhou University,Yangzhou 225000,China)
出处
《飞行力学》
CSCD
北大核心
2023年第5期23-29,共7页
Flight Dynamics
基金
国家自然科学基金资助(62003162)
江苏省自然科学基金资助(BK20200416)
空间智能控制技术实验室开放基金资助(HTKJ2022KL502015)
中国高校产学研创新基金资助(2021ZYA02005)。
关键词
飞行性能
过载能力
长短期记忆
性能预测
flight performance
overload capacity
LSTM
performance prediction
