基于自适应高斯-牛顿迭代三步延迟时空融合的无人机测高

Altitude measurement of UAV based on adaptive Gauss-Newton iterative three-step delay space-time fusion

针对无人机巡航测高过程中,采用单一传感器进行高度测量时受到环境干扰较大且精度较差,而现有多传感器融合方法精度较低的问题,提出一种基于自适应高斯-牛顿迭代三步延迟时空融合的无人机测高方法.首先,将优化后的三步延迟时空融合算法作为第一层融合,有效避免了传感器出现故障所导致的测量误差;其次,采用加速度计对融合后的数据进行检测和辅助测高,将基于参数迭代的自适应高斯-牛顿滤波器作为第二层融合模型算法,可以在保证稳定性的前提下,提高无人机巡航或悬停时的精度.最后,采用实际高度测量数据进行实验.实验结果表明,与二步延迟融合方法比较,最大误差减小25.4%,均方根误差减小26.4%.同时,该测高方法简化了相关量测方程和状态量维数,降低了计算量,系统结构更加简洁,有利于实际应用.

Aiming at the problem that the height measurement process of the UAV during cruising,the use of a single sensor for height measurement is subject to large environmental interference and poor accuracy, while the existing multi-sensor fusion methods have low accuracy. A UAV altimetry method adapted to Gauss-Newton iterative three-step delayed space-time fusion. First, the optimized three-step delay space-time fusion algorithm is used as the first layer of fusion, which effectively avoids the measurement error caused by the failure of the sensor. Second,the accelerometer is used to detect the fused data and assist in height measurement,and The adaptive Gauss-Newton filter based on parameter iteration is used as the second-layer fusion model algorithm, which can improve the accuracy of the UAV when cruising or hovering under the premise of ensuring stability. Finally, experiments are carried out using data from actual height measurements. The results show that compared with the two-step delay fusion method, the maximum error is reduced by 25.4%, and the root mean square error is reduced by 26.4%. At the same time,the altimetry method simplifies the related measurement equations and the dimension of the state quantity, reduces the calculation amount, and makes the system structure more concise, which is beneficial to practical application.