多源多层的融合自适应加权积分旋翼无人机测高算法

Multi-source Multi-layer Fusion Adaptive Weighted Integral Rotor UAV Altimetry Algorithm

针对目前在无人机测高过程中,仅利用单一传感器对高度进行测量会导致结果精度差且受扰动影响较大,以及现有基于多传感器融合算法系统提升精度有限的问题,提出了一种多源多层的融合自适应加权积分旋翼无人机测高算法。首先,在设计相关认知支持度的基础上,结合测量获取的数据进行第一层融合,提出基于一种邻近相关认知一致性样本均值和协方差相关指数度量的融合权值分配算法,使得高度误差低于单个传感器误差,提升精度;然后,在第二层融合设计了自适应互补滤波方法,对高度辅助测量时结合加速度计数据,进一步减小误差,确保无人机在巡航和工作过程中对高度的控制能更准确和稳定。仿真实验结果表明,该系统相对于传统的利用单一传感器测高算法,均方根误差减小了51.9%,最大误差减小了59.8%,且无需过多的先验知识和噪声特性,效果十分显著。同时,该算法简单易行,数据量小,在工程方面易于实现。

At present,in the process of Unmanned Aerial Vehicle(UAV)height measurement,only using a single sensor to measure the height will lead to poor accuracy and be greatly affected by disturbance,and the existing system based on multi-sensor fusion algorithm has limited accuracy improvement.This paper proposes a multisource multi-layer fusion adaptive weighted integral rotor UAV height measurement algorithm.Firstly,based on the design of relevance cognitive support and the first level fusion of the measured data,a fusion weight allocation algorithm based on the measurement of the sample mean and covariance correlation index is proposed,which makes the height error lower than the single sensor error,and improves the accuracy;Then,the adaptive complementary filtering method is designed in the second level fusion,which combines the accelerometer data to further reduce the error,and ensure that the UAV can control the altitude more accurately and stably in the process of cruise and work.The simulation results show that compared with the traditional single sensor algorithm,the root mean square error of the system is reduced by 51.9%,and the maximum error is reduced by 59.8%.At the same time,the algorithm is simple and easy to implement in engineering.