飞行实测数据气动力建模研究  

Research on Aerodynamic Modeling Based on Rough Set and Wavelet Neural Network from Flight Measured Data

作　　者：李永新[1] 甘旭升[2] 屈虹[2] 赵海涛[2] 

机构地区：[1]西京学院,陕西西安710123 [2]空军工程大学空管领航学院,西安710051

出　　处：《计算机仿真》2016年第8期72-75,共4页Computer Simulation

摘　　要：在飞行器气动性能优化建模中,准确的气动力模型是飞行器控制律设计的基础。针对建模方法不当以及模型结构不合理难以有效提高建模精度的问题,提出了一种粗糙集与小波神经网络(WNN)的气动力建模方法。利用WNN良好的非线性建模能力,在飞行实测数据基础上构建飞行器的气动力模型,并通过一种启发式的粗糙集约简算法优化待定的WNN结构,以改善模型性能。横侧向与纵向气动力建模仿真结果表明,与经验公式相比,粗糙集约简确定的WNN,网络结构简化合理,模型性能有效提升,预测输出误差可控制在0.005以内。证明为飞行器气动力性能优化建模提供了依据。The accurate aerodynamic model is the basis of flight simulation and control law design for aircraft. For the difficulties in improvement of modeling accuracy because of improper modeling method and unreasonable model structure,an aerodynamic modeling method is proposed based on rough set and Wavelet Neural Network（ WNN）.By using the good nonlinear modeling ability of WNN,the aerodynamic model of aircraft is established based on flight measured data,and a heuristic rough set reduction algorithm is used to optimize WNN structure,in order to improve the model performance. Lateral and longitudinal aerodynamic modeling simulation results show that,compared with the empirical formula,rough set can determine a more simplifying and reasonable network structure for WNN with performance improvement,and the prediction output error can be controlled within 0. 005. It is proved that the proposed method is an effective and feasible method for aerodynamic modeling of aircraft.

关 键 词：气动力建模 小波神经网络 粗糙集 飞行器 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构] TP183[自动化与计算机技术—计算机科学与技术]