论文部分内容阅读
飞行动力学辨识算法的一个关键问题是,如何通过简单的机动获取所关心频率范围的响应特性。短时倍脉冲是一种易于实施的激励信号,兼顾试飞安全性与经济性,但与频域辨识法通常使用的扫频输入激励相比,短时机动频谱范围窄、信噪比低,一般难以得到准确的辨识结果。对如何基于短时机动飞行试验数据,提高辨识结果准确性的问题进行了研究。首先分析了经典Welch谱估计进行时域-频域转换过程中,影响非参数模型辨识精度的主要因素,提出了削减窗函数边缘缩减效应的数据预处理方法,并结合多窗口综合技术,提高频域特性辨识结果的精度。在参数化模型辨识过程中,针对有限频谱范围,提出了利用相干函数和功率谱密度加权综合,确定等效拟配的频率范围和频率节点的自适应方法,使得低阶等效拟配与输入激励信号高度相关,提高参数化模型辨识的精度、一致性和适应性。通过不同类型飞机的大量短时机动和少量扫频飞行试验数据模型辨识的工程应用示例,验证了动力学辨识优化方法算法稳定、结果准确,可满足飞行品质模态特性评价等应用需求。
A key issue of the aerodynamic recognition algorithm is how to obtain the response characteristics of the frequency range of interest by simple maneuvering. Short time multiplier is an easy-to-implement excitation signal that takes into account the safety and economy of test flight. However, compared with the commonly used sweep input excitation in frequency domain, short-time maneuvering has narrow spectrum range and low signal-to-noise ratio. Difficult to get accurate identification results. How to improve the accuracy of the recognition results based on the experimental data of short-term maneuver flight was studied. Firstly, the main factors affecting the identification accuracy of nonparametric models in the time-frequency domain transformation of classical Welch spectrum estimation are analyzed. A data preprocessing method of reducing the edge reduction effect of window functions is proposed. Combined with the multi-window synthesis technique, Accuracy of domain character recognition results. In the process of parametric model identification, aiming at the limited spectrum range, an adaptive method to determine equivalent frequency range and frequency nodes with coherent function and power spectral density weighted synthesis is proposed. Incentive signals are highly correlated and improve the accuracy, consistency and adaptability of parametric model identification. Through a large number of examples of engineering applications for the identification of a large number of short-term maneuvers and a few flight-type flight test data models of different types of aircraft, the method of dynamic identification optimization is verified to be stable and the results are accurate to meet the application requirements of flight quality modal characteristics evaluation.