基于BP神经网络的气象威胁度建模和评估方法研究  

Weather Threat Modeling and Assessment Based on BP Neural Networks

作　　者：朱国涛[1,2] 周树道[1] 叶松[1] 王敏[1] 王彦杰[1] 

机构地区：[1]中国人民解放军理工大学气象学院 [2]中国人民解放军94195部队

出　　处：《电光与控制》2011年第3期69-71,90,共4页Electronics Optics & Control

基　　金：国家自然科学基金(40976062)

摘　　要：对执行非作战任务的无人机航迹规划而言,气象要素是必须考虑的主要因素之一。重点针对气象要素对无人机航迹规划的影响进行分析,给出了8种严重影响飞行安全的气象相关因子。结合BP神经网络算法建立了气象威胁度模型,并对气象威胁度进行了仿真评估分析。仿真结果表明,该气象威胁度模型简单实用,评估方法具有良好的效用,能较迅速、准确地反映实际的气象威胁度等级。For those UAVs to execute non-combat tasks,weather is one of the important factors for route planning which should be taken into consideration.The influence of weather to UAV route planning was analyzed,and eight weather related factors were proposed which have great effect on flight safety.BP neural network algorithm was used for establishing the model of weather threat about these weather related factors,and the weather threat level was assessed based on the model.Simulation results showed that the weather threat model and the assessment method are simple and effective,which can reflect the real weather threat level rapidly and exactly.

关 键 词：航迹规划 威胁评估 神经网络 气象 BP算法 

分 类 号：V271.4[航空宇航科学与技术—飞行器设计] TP391.9[自动化与计算机技术—计算机应用技术]