基于改进Zoom – FFT的无人机仿地飞行毫米波雷达测高方法

Altitude Measurement Method for UAV Terrain-following Flight Using Millimeter-wave Radar Based on Improved Zoom – FFT

摘要:

为解决频率估计精度受限于频率分辨率和栅栏效应的难题,以提高无人机测高精度,提出了一种基于改进Zoom – FFT算法的毫米波雷达高精度频率估计方法。在建立Zoom – FFT算法误差模型、系统分析误差关键因素的基础上,提出了一种基于频率修正的改进Zoom – FFT频率估计方法:首先,使用FFT算法对中频信号进行频率粗估计,确定频率的感兴趣区间;然后,通过Zoom – FFT频谱细化技术,提高感兴趣频带内的频率分辨率;最后,提出二次多项式拟合细化频谱的频率修正方法,拟合最大谱线及其相邻谱线,求解修正频率,以有效补偿栅栏效应带来的偏差。仿真表明,本文方法在不同信噪比和频率条件下,均表现出良好的频率估计精度与稳定性,性能比传统方法更接近理论极限。进一步的无人机测高实验表明,在2 m悬停状态下,本文算法的无人机测高最大误差为2. 8 mm,实现了毫米级精度;在飞行状态下,测高与测角精度也优于传统Zoom – FFT等算法,测量精度提高33%以上,证明了本文方法的有效性与实用性。

Abstract:

Aiming to solve the problem that frequency estimation accuracy is limited by frequency resolution and the fence effect, and improve the altitude measurement accuracy of unmanned aerial vehicles (UAVs), a high-precision frequency estimation method was proposed for millimeter-wave radar based on an improved Zoom – FFT algorithm. Based on the establishment of an error model of the Zoom – FFT algorithm and a systematic analysis of the key error factors, an improved Zoom – FFT frequency estimation method using frequency correction was proposed. Firstly, the fast Fourier transform (FFT) algorithm was used to perform coarse frequency estimation on the intermediate frequency (IF) signal to determine the frequency band of interest. Secondly, the Zoom – FFT spectrum refinement technique was applied to enhance the frequency resolution within the band of interest. Finally, a frequency correction method based on quadratic polynomial fitting of the refined spectrum was proposed, which fit the maximum spectral line and its adjacent spectral lines to obtain the corrected frequency, thereby effectively compensating for the bias caused by the fence effect. Simulations demonstrated that the proposed method exhibited good frequency estimation accuracy and stability under different signal-to-noise ratio and frequency conditions, and its performance was closer to the theoretical limit than that of conventional methods. Further UAV altitude measurement experiments showed that in a hovering state at 2 m altitude, the maximum altitude measurement error of the proposed algorithm was 2. 8 mm, achieving millimeter-level accuracy. In the flying state, both altitude and angle measurement accuracies were superior to those of the traditional Zoom – FFT and other algorithms, and the measurement accuracy was improved by more than 33% , proving the effectiveness and practicality of the proposed method.