引用本文: | 胡爽,朱纪洪.基于飞行数据多重分区的高机动飞机气动建模及参数辨识[J].控制理论与应用,2016,33(10):1289~1295.[点击复制] |
HU Shuang,ZHU Ji-hong.Multiple partitioning of flight data for manoeuvrable aircraft aerodynamic modeling and parameter estimation[J].Control Theory and Technology,2016,33(10):1289~1295.[点击复制] |
|
基于飞行数据多重分区的高机动飞机气动建模及参数辨识 |
Multiple partitioning of flight data for manoeuvrable aircraft aerodynamic modeling and parameter estimation |
摘要点击 3656 全文点击 1935 投稿时间:2016-04-27 修订日期:2016-09-30 |
查看全文 查看/发表评论 下载PDF阅读器 |
DOI编号 10.7641/CTA.2016.60260 |
2016,33(10):1289-1295 |
中文关键词 空气动力 气动建模 参数辨识 多重分区 |
英文关键词 aerodynamics aerodynamic modeling parameter estimation multiple partitioning |
基金项目 国家“973”计划项目(2012613189)资助. |
|
中文摘要 |
在准定常假设下, 飞机在大迎角或大幅机动飞行时, 其气动特性呈现非线性特点. 常用基于配平状态下小
幅机动飞行辨识所得的线性气动模型已不再适用. 为解决这一问题, 提出一种飞行数据多重分区方法, 通过各区间
的局部线性化以表征气动特性的全局非线性. 各区间中, 针对气动力和力矩系数的静态项、动导数项及控制导数项
进行泰勒级数展开, 提出一种通用气动模型, 并利用最小二乘类方法辨识各项气动参数. 根据某现代战斗机仿真飞
行试验数据, 辨识相关气动参数并与真实值进行比较, 结果表明两者吻合较好. 试验结果验证了所述飞行数据多重
分区方法和通用气动模型的有效性. |
英文摘要 |
Under the quasi-steady assumption, aerodynamic characteristics show marked nonlinearity in high angle of
attack or large amplitude maneuvers of manoeuvrable aircraft. A linear aerodynamic model generally identified from small
amplitude maneuvers about trim conditions is no longer suitable for this case. In order to solve this problem, a method
of multiple partitioning of flight data is proposed, which can characterize aerodynamic global nonlinearity by partitioned
linearity. In each partitioned subset, a general aerodynamic model is formed by the Taylor’s series expansion of static terms,
dynamic stability derivatives and control derivatives of aerodynamic force and moment coefficients. The parameters in this
aerodynamic model are estimated by using the class of least squares methods. According to the simulation flight test data
of a modern fighter aircraft, the aerodynamic parameters are estimated which fit well with the true values. The test results
verify the proposed multiple partitioning of flight data and the general aerodynamic model. |