一种涵道无人直升机模型参数辨识试验策略

An Identification Method of Parameter for Ducted Fan Unmanned Helicopter Model

针对样例涵道无人直升机,考虑机体振动等环境因素对输出数据的污染,提出一种样例无人直升机悬停模态下模型参数辨识方法.针对带角速率反馈的广义直升机参数辨识模型,利用直升机角速率、线加速度和操纵量依次独立辨识纵向通道模型参数Iy,Ib和CTv.首先根据传感器输出噪声方差先验信息筛选观测数据,有效剔除由此产生的部分野值样本,然后进一步采用低通滤波器对观测数据进行滤波,通过最小二乘法进行二次辨识.试验结果表明,该方法可显著提高模型参数辨识精度并降低参数估计的标准差.

For a prototype-ducted fan unmanned helicopter, considering the contaminated data effects of the mechani- cal vibration from the airframe, the method of parameter identification of unmanned helicopter hovering model is pro- posed. Based on generalized helicopter parameter model with angular velocity feedback, longitudinal channel model parameters/, Ib and Cr are identified by angular rate, linear acceleration and control input successively. Firstly, the screening of the observed data according to prior information of the sensor output noise variance was performed, which rejected the abnormal sample. Secondly, the observed data were filtered by low-pass filter. Finally, the sec- ondary identification was made by the least squares method. The experimental results showed that this method could improve the precision and reduce the standard deviation of parameter identification significantly.