基于无人机航向的不规则区域作业航线规划算法与验证

为尽可能地减少飞行总距离和多余覆盖面积，节省无人机的能耗和药液消耗，研究了一种基于作业方向的不规则区域作业航线规划算法。该算法根据指定的作业方向，可快速规划出较优的作业航线，也可在未指定作业方向的情况下，给出某一推荐的作业方向与航线，使整个作业过程能耗和药耗最优。仿真结果表明，在一块面积为983.125 m2的不规则凸五边形作业区域内，采用该算法进行航线规划，无人机作业的多余覆盖率最低可达到11.5%，而且作业面积越大，优化效果越明显，在同样的地块进行田间试验，得到最低多余覆盖率为2.8%，证明了该算法的可行性。该研究可为自主作业无人机的航线规划算法提供参考。

Route planning algorithm and verification based on UAV operation path angle in irregular area

Abstract: Route planning is an important step for the automatic unmanned aerial vehicles, and the quality of route planning will directly affect the energy consumption and pesticide consumption of unmanned aerial vehicles. Currently plant protection unmanned aerial vehicles rely too much on artificial remote, the flight routes are not planned precisely and the actual flight routes are often seriously different from the theoretical routes, especially in the irregular areas. So research on route planning method suitable for automatic unmanned aerial vehicles is necessary. Aimed to the operation area with irregular convex polygon shape, a kind of operation route planning algorithm is proposed to extract the unmanned aerial vehicles' operation routes in any specified operating direction quickly. Latitude and longitude coordinates of border points are converted to the metric ones, environment coordinate is built to convert operation area borders to coordinate volume, and coordinate transformation is according to the operation direction, which makes it easier to compute. The operation area is divided into a plurality of sub-regions, and each sub-region's start and stop operation waypoint are set by considering the flight distance and pesticide consumption; all operation waypoints are connected to get operation routes, which can minimize the excess coverage area and the waste of pesticides. When the operation direction is not specified, the route planning algorithm can also give a recommended operation direction and route to make energy and pesticide consumption minimal in the entire operation process. The simulation results showed that using this algorithm in an irregular convex pentagon area of 983.125 m2, when the course angle of the operation routes was set to be 0o, 45o, 90o and 135o, corresponding total flight distance was 273.38, 291.30, 273.68 and 293.78 m, coverage area was 1 121.8, 1 195.5, 1 169.2 and 1 197 m, excess coverage rate was 14.1%, 21.6%, 18.9% and 21.8%, respectively. The total flight distance and coverage area with course angle of 0o were the least among 4 kinds of course angles, and its energy and pesticide consumption were also the least. While the best operating course angle of the convex pentagon area was 100o, and its corresponding coverage area and excess coverage rate were 1 096.5 m2and 11.5%. In field tests, the minimum excess coverage rate was 2.8%, which showed the feasibility of the operation route planning algorithm; while the test results also showed the deviations between theoretical and practical operation routes, and the possible reasons were GPS positioning error, wind and unstable center of gravity due to liquid, so the GPS positioning accuracy needed to be improved, and flight control systems also needed further improvement to make the flight attitude of unmanned aerial vehicle more stable. Operation routes are planned by the route planning algorithm before operation, and energy and pesticide consumption can be previously estimated, which can save the time consumed by preparation work and make unmanned aerial vehicle operation more efficient. Nowadays, unmanned plant protection operation by unmanned aerial vehicle becomes popular, and planning and management of the operation also become very important, and hence this flight route planning algorithm not only saves manpower required by route planning, but also makes operation management easier. In addition, the algorithm also reduces the deposition of pesticides in non-operating areas and is conducive to environment protection. The algorithm is suitable for automatic unmanned aerial vehicles and can be widely used in the area of precision agriculture.