基于自适应AR模型巡航飞行参数预测研究  

CRUISE FLIGHT PARAMETERS PREDICTION BASED ON ADAPTIVE AR MODEL

作　　者：钱宇[1] 王立新 张恒 刘瑜 Qian Yu;Wang Lixin;Zhang Heng;Liu Yu(Civil Aviation Flight University of China,Guanghan 618307,Sichuan,China)

机构地区：[1]中国民用航空飞行学院,四川广汉618307

出　　处：《计算机应用与软件》2024年第4期73-79,共7页Computer Applications and Software

基　　金：国家自然科学基金民航联合基金项目(U1833201);四川省教育厅自然科学基金项目(16ZA0021)。

摘　　要：为更准确实现飞行参数趋势预测,提出一种基于自适应自回归(AR)模型的稳定巡航飞行参数预测方法。根据稳定巡航参数筛选条件,获取建模所需飞行参数。利用卡尔曼滤波原理估计AR模型参数,并与飞行参数构建系统方程,利用无迹卡尔曼滤波实时更新、修正AR模型参数估计值,将自适应AR模型的预测值与曲线拟合模型和灰色模型的预测值进行对比。以波音B777-300ER飞机的快速存取记录器数据样本进行仿真验证,结果表明:自适应AR模型在数据预测和收敛速率方面均更优,可有效降低预报模型随步数增加导致的精度误差,提高参数预测准确性。研究在飞机维修保障、状态监控与预测等方面具有重要作用。In order to realize the trend prediction of flight parameters more accurately,a stable cruise flight parameter prediction method based on adaptive auto regressive(AR)model is proposed.According to the screening conditions of stable cruise parameters,the flight parameters required for modeling were obtained.The parameters of AR model were estimated by Kalman filter principle,and the system equations were constructed with flight parameters.The parameters of AR model were updated and modified by unscented Kalman filter(UKF)in real time.The predicted values of adaptive AR model were compared with those of curve fitting model and grey model.The data samples of Boeing B777-300ER quick aircraft recorder(QAR)were used for simulation verification.The results show that the adaptive AR model is better in data prediction accuracy and convergence rate,which can effectively reduce the accuracy error of prediction model with the increase of steps and improve the accuracy of parameter prediction.This research is of great significance in aircraft maintenance support,condition monitoring and prediction.

关 键 词：无迹卡尔曼滤波 自适应AR模型 飞行参数预测 曲线拟合模型 灰色模型 

分 类 号：TP206[自动化与计算机技术—检测技术与自动化装置] TP3[自动化与计算机技术—控制科学与工程]