论文部分内容阅读
在输入-输出数据中带有噪声时,传统的最小二乘辨识算法给出模型参数的有偏估计。当噪声方差的估计值可获得时,采用偏差补偿算法能够得到模型参数的一致性估计。在辅助变量算法的基础上结合偏差补偿算法进行推广得到偏差补偿辅助变量辨识算法。采用适用于噪声环境的偏差补偿辅助变量辨识算法,可准确地辨识飞机的颤振模态参数,该算法结合传递函数模型,将带噪声系统的辨识问题转化为迭代求解问题,用来解决输入噪声为白噪声,而输出噪声为有色噪声的复杂辨识情况。利用该算法可将噪声的方差值和传递函数中的模型参数迭代地估计出来。最后利用试飞试验数据辨识飞机颤振的系统参数,将算法与经典的辅助变量算法进行比较,验证了该方法的有效性。
In the presence of noise in the input-output data, the traditional least-squares identification algorithm gives a biased estimate of the model parameters. When the estimated noise variance is available, a bias compensation algorithm can be used to obtain consistent estimates of model parameters. Based on the auxiliary variable algorithm, the deviation compensation algorithm is used to derive the offset compensation auxiliary variable identification algorithm. The algorithm of offset compensation for auxiliary variable is applied to the noise environment to identify the flutter modal parameters accurately. The algorithm combined with the transfer function model transforms the identification of the noisy system into an iterative solution to the problem of input noise Is white noise, and the output noise is a complex identification of colored noise. The algorithm can be used to estimate the variance of noise and the model parameters in the transfer function iteratively. Finally, the system parameters of aircraft flutter are identified by using the test flight test data. The algorithm is compared with the classical auxiliary variable algorithm to verify the effectiveness of the proposed method.