摘要
飞机在地面的结霜会破坏飞机的气动外形,对飞机的飞行性能产生不利的影响。为了对地面飞机结霜进行准确监测与预测,利用多种传感器对机翼的结霜环境进行实验研究,通过分析不同环境下模拟机翼的霜层生长曲线,探寻和统计飞机机翼的地面结霜规律。对结霜实验的数据进行分析,提出了不同结霜环境下的多项式回归预测模型。同时将多项式进行正交化,避免了多重共线性,使预测模型的曲线拟合程度高,预测值具有较高的精度,实现了对飞机的地面结霜进行快速精确的预测。
The frosting of aircraft on the ground destroys aerodynamic shape and has a negative effect on the flight performance. In order to monitor and forecast on aircraft frosting on the ground accurately,experimental research on airfoil frosting is carried out using variety of sensors; by analyzing frost layer growth curve of simulating airfoil in different environment conditions,law of airfoil frosting is researched. Analyze datas of frosting experiment,propose polynomial regression predicting model in different frosting environments. Meanwhile,multicollinearity is avoided by orthogonalizing polynomials,which make prediction model has high curve fitting and high precision,fast and accurate prediction for ground frosting of aircrafts is realized.
出处
《传感器与微系统》
CSCD
2015年第2期40-43,共4页
Transducer and Microsystem Technologies
基金
国家自然科学基金资助项目(60939001)
国家科技支撑计划资助项目(2012BAG04B02)
民航局科技基金资助项目(60932016)
关键词
结霜
光纤传感器
多项式
回归分析
预测
frosting
fibre-optic sensor
polynomial
regression analysis
prediction
