论文部分内容阅读
在线辨识在现代飞行控制系统设计中扮演越来越重要的角色,飞行器模型的在线更新使得人们可以采用更智能的控制方法。基于计算精度和速度的考虑,在线辨识方法通常以递推方式进行,主要分为时域和频域两大类方法。在建立飞行器系统模型结构的基础上,利用频域递推傅里叶变换及最小二乘方法,实现对气动及控制偏导数的在线辨识。针对某飞机纵向通道的在线辨识仿真验证该方法有效,且计算速度和收敛速度快,辨识结果与参数真实值之间的一致性好,方法对传感器噪声有较强的适应性。最后,分析比较了各种典型激励信号对辨识结果的影响,为进行实际在线辨识试验提供了参考依据。
Online identification plays an increasingly important role in the design of modern flight control systems. The online updating of aircraft models allows one to adopt more intelligent control methods. Based on the calculation accuracy and speed considerations, online identification methods are usually recursive approach, mainly divided into time domain and frequency domain two categories of methods. Based on the establishment of the aircraft system model structure, the on-line identification of aerodynamic and control partial derivatives is realized by using the frequency domain recursive Fourier transform and the least square method. This method is validated by on-line identification of a longitudinal channel of an airplane. The proposed method is effective and fast in calculation and convergence. The consistency between the identification result and the true value of the parameter is good. The method has strong adaptability to sensor noise. Finally, the influence of various typical excitation signals on the identification results is analyzed and compared, which provides a reference for the actual on-line identification test.