基于组合导航与EKPF飞行器的地形边界与面积估计

Estimation of Boundary and Area Using Aircraft with Integrated Navigation and EKPF

提出了一种使用四旋翼飞行器作为测量平台的地形边界与面积估计算法.该算法采用紧耦合闭环组合导航系统(捷联惯性导航系统与全球卫星定位系统)获取待测地形边界点的定位数据,分别使用Pauta准则与扩展卡尔曼粒子滤波方法剔除异常数据与处理定位数据.算法在得到的最终定位数据的基础上,估算待测地形与面积.飞行器使用前视和下视摄像头确定飞行方向和选择边界点.使用该算法可以实现不规则的凸多边形、凹多边形和弧段等地形的精确估计,估计误差可以在±1.2%以内.实际飞行实验结果验证了该算法的可行性和准确性.

An algorithm was proposed to estimate the boundary and area of terrains with the quadrotor aircraft as the measuring instrument. The tightly coupled closed-loop integrated navigation system （strap-down inertial navigation system and global positioning system） was used to collect the positioning data. The Pauta criterion and EKPF （extended Kalman particle filter） were used to exclude the anomalous positioning data and dispose the remaining positioning data respectively. The boundary and area were estimated using the proposed algorithm based on the final positioning data. The front facing and bottom facing cameras of the aircraft were used to determine the flight direction and select the boundary points. This algorithm could be used to estimate irregular convex polygons, concave polygons and segmental arcs. The estimation error was limited within ±1.2%. The actual flight test results indicated the feasibility and accuracy of the proposed algorithm.