【摘要】 无人机协同区域覆盖在军事侦察、植保等领域广泛应用，然而现有算法仍然存在区域覆盖重复率高、规划轨迹方向变化频繁等问题。为此，提出基于改进牛耕法和惯性轨迹生成的多机分布式区域覆盖算法。算法基于地图列方向划分栅格地图，在牛耕法的基础上分裂线性结构并更新线性结构集合，从而实现低重复率覆盖邻近区域。此外，在广度优先算法的基础上，建立了一种考虑轨迹惯性变化率代价的模型，进而生成了低轨迹方向变化率的惯性轨迹。仿真和实验结果表明，基于改进牛耕法和惯性轨迹生成的区域覆盖算法相对于牛耕法区域，重复覆盖率降低3%,规划轨迹的惯性变化率相对于常用的广度优先算法能够显著降低40%。更多还原

【Abstract】 Area coverage of UAV coordination has been widely used in military investigation, pesticide spraying and other fields in the field of UAV. However, the existing methods suffered from high repetition rate, lots of turns in planned paths. To address these problems, a multi-agent distributed area coverage algorithm based on improved boustrophedon algorithm and inertial trajectory generation algorithm is proposed. The algorithm divides the grid map based on the map column direction, splits the linear structure and updates the linear structure set based on the boustrophedon algorithm in order to cover the adjacent area with low repetition rate. Based on the breadth first algorithm, the algorithm establishes a model considering the cost of trajectory inertial change rate, and generates the inertial trajectory with low trajectory direction change rate. The simulation and experimental results show that the area repeated coverage rate of the algorithm is 3% lower than that of boustrophedon algorithm, and the inertial change rate of planned trajectory can be significantly reduced by 40% compared with the commonly used breadth-first search algorithm.更多还原