基于粒子滤波的无人机自主轨迹视觉导航控制方法研究

针对现有无人机导航控制方法存在的控制效果不佳的问题,本文提出一种基于粒子滤波的无人机自主轨迹视觉导航控制方法研究。利用粒子滤波算法,实现对无人机自主轨迹视觉导航控制方法的优化设计。采用栅格法构建无人机飞行环境地图,根据无人机的机械组成结构和工作原理,构建运动状态模型。利用内置的摄像机设备采集视觉图像,执行图像灰度转换、几何校正、滤波等预处理步骤。通过对视觉图像的特征提取,判断当前环境是否存在障碍物。利用粒子滤波算法确定无人机位姿,结合障碍物识别结果规划无人机的自主飞行轨迹。将位置、速度和姿态角的控制量计算结果,输入到安装的导航控制器中,完成无人机的自主轨迹视觉导航控制任务。通过实测分析得出结论：应用设计的导航控制方法,其位置误差、速度误差以及姿态角误差均维持在预设值以下,即设计的导航控制方法具有良好的控制效果。

Research on visual navigation control method of UAV autonomous trajectory based on particle filter

Aiming at the problem of poor control effect of existing UAV navigation control methods, a visual navigation control method of UAV autonomous trajectory based on particle filter is proposed in this paper. The particle filter algorithm is used to optimize the visual navigation control method of UAV autonomous trajectory. The flight environment map of UAV is constructed by grid method, and the motion state model is constructed according to the mechanical composition, structure and working principle of UAV. The built-in camera equipment is used to collect visual images and perform preprocessing steps such as image gray conversion, geometric correction and filtering. Through the feature extraction of visual image, it can judge whether there are obstacles in the current environment. The position and attitude of UAV are determined by particle filter algorithm, and the autonomous flight trajectory of UAV is planned combined with the obstacle recognition results. The calculation results of position, speed and attitude angle are input into the installed navigation controller to complete the autonomous trajectory visual navigation control task of UAV. Through the actual measurement and analysis, it is concluded that the position error, velocity error and attitude angle error of the designed navigation control method are maintained below the preset value, that is, the designed navigation control method has good control effect.