基于RBF神经网络与LS-SVM方法的无人机飞行数据预处理方法

Research on flight data pretreatment of UAV based on the method of RBF neural network and LS-SVM

在对无人机飞行数据预处理方法分析比较的基础上,将RBF神经网络和最小二乘支持向量机相结合,提出1种飞行数据预处理的组合方法,进行优势互补,对单一方法的处理结果进行加权平均。仿真结果表明,该方法增强了算法的可靠性,提高了数据预处理的精度,具有较强的可操作性和适用性。

On the basic of analyzing and comparing the flight data pretreatment methods of UAV,a novel combined method in flight data pretreatment is proposed.This method combines the advantages of RBF neural and Least Squares Support Vector Machines（LS-SVM）,and processes the results of single method with weighted mean.The simulation result shows that the reliability of the method is strengthened,the precision in flight data pretreatment is enhanced,and the maneuverability and applicability of the method are good.