-
如图 1所示,设置杆塔塔高为h,无人机飞行高度hw,传感器测量最大测量距离为d,安全冗余距离为s,传感器与地面距离dw,无人机与杆塔连线水平面上垂直距离为w。为保证无人机飞行安全与测量有效性,应该满足以下约束条件:
$ {d - s > {h_{\rm{w}}} > h + s} $
(1) $ {w < \left( {1 - \frac{h}{{{h_{\rm{w}}}}}} \right)\sqrt {{d^2} - {h_{\rm{w}}}^2} } $
(2) $ {{d_{\rm{w}}} < d} $
(3) 飞行约束空间并不是静态的,而是随着无人机与山区环境的相对高度而变化。飞行约束环境核心是求取出无人机在此位置下与山区环境地面的相对高度,进而利用上述公式求取飞行约束环境。为了保证无人机飞行轨迹稳定,本文中设置多旋翼无人机飞行相对高度稳定为hw。
约束空间建模核心在于使无人机激光扫描区域与杆塔检测区域相重合,需要考虑测量距离、飞行高度、横向距离等因素,构建优化约束条件如前面所示。其中测量地面目标最大距离dw限定小于传感器测量最大有效距离d,保证激光扫描能够收集到地面信息。无人机相对飞行高度hw保持定值,避免无人机飞行过低撞上杆塔的情况。无人机横向距离w小于当前飞行高度对应的最大横向距离,限制无人机飞行的偏移量以优化路径航程大小。
-
激光传感器的测量角度范围、测量距离、激光波长等参量的不同,无人机搭载激光传感器的测量范围与效果也不同[24],存在一个有效测量区间。进行无人机激光扫描路径规划,需要保证激光传感器有效测量区间与电力走廊空间相重合。因此,本文中研究激光传感器最大测量距离d与测量角度θ等特性,定义无人机与垂直距离w、无人机测量杆塔的最大距离R作为规划路径限制指标(如图 2所示)。
传感器最大测量距离d限制无人机与杆塔的垂直距离w的范围。本文中设置(Sx, Sy), (Gx, Gy), (Nx, Ny)分别为两个杆塔和无人机的xy轴坐标。如图 3所示,首先求取无人机与两个杆塔之间的水平面上的距离l1, l2,以及两个杆塔之间的距离l3。利用3个边的关系可以求得无人机与两个杆塔连线的垂直距离w:
$ {{l_1} = \sqrt {{{({N_x} - {S_x})}^2} + {{({N_y} - {S_y})}^2}} } $
(4) $ {{l_2} = \sqrt {{{({N_x} - {G_x})}^2} + {{({N_y} - {G_y})}^2}} } $
(5) $ {{l_3} = \sqrt {{{({S_x} - {G_x})}^2} + {{({S_y} - {G_y})}^2}} } $
(6) $ {w = \sqrt {l_2^2 - {{\left( {\frac{{l_2^2 + l_3^2 - l_1^2}}{{2{l_3}}}} \right)}^2}} } $
(7) 传感器测量角度θ与最大测量距离d限制无人机测量杆塔的最大距离R的大小,如图 4所示。设置无人机飞行高度为hw、无人机测量杆塔的最大距离R与传感器测量角度θ、最大测量距离d关系如下:
$ R = \left\{ {\begin{array}{*{20}{l}} {(1 - \frac{h}{{{h_{\rm{w}}}}})\sqrt {{d^2} - {h_{\rm{w}}}^2} ,\left( {\frac{{{h_{\rm{w}}}}}{{{\rm{cos}}\theta }} > d} \right)}\\ {\frac{{{h_{\rm{w}}} - h}}{{{\rm{tan}}\theta }},\left( {\frac{{{h_{\rm{w}}}}}{{{\rm{cos}}\theta }} \le d} \right)} \end{array}} \right. $
(8) 将无人机测量杆塔的最大距离R作为电线杆塔吸引域的半径,无人机到达吸引域中任意一点均可保证杆塔检测有效性,如图 5所示。
本文中研究Velodyne VLP-16, RS-LiDAR-16, HDL-32E 3种激光传感器的特性,进行路径规划。激光传感器如图 6所示, 相关参量如表 1所示。
Table 1. Laser sensor parameters
model d/m θ Velodyne VLP-16 100 ±15°, vertical RS-LiDAR-16 150 ±15°, vertical HDL-32E 100 +10°~-30°, vertical -
无人机路径规划需要考虑无人机的飞行航程问题。为保证无人机安全执行巡检任务,无人机需要在一定时间内完成任务并返航。定义最大航程lmax,进行无人机路径规划时计算前往下一个目标点并开始返航的航程是否超限。如果超限,规划无人机从此刻目标点立即返航,否则继续规划前往下一个目标点[25]。规划轨迹li与最大航程关系如下所示:
$ \sum {{l_i}} + {l_{{\rm{start}}}} < {l_{{\rm{max}}}} $
(9) $ \begin{array}{*{20}{c}} {{l_{{\rm{ start }}}} = }\\ {\sqrt {{{({x_i} - {x_{\rm{s}}})}^2} + {{({y_i} - {y_{\rm{s}}})}^2} + {{({z_i} - {z_{\rm{s}}})}^2}} } \end{array} $
(10) 式中,(xi, yi, zi)为当前无人机位置坐标,(xs, ys, zs)为无人机路径规划起始点。
-
路径代价函数是无人机路径规划的重要评价指标,作用是保证无人机飞行的高效性与安全性。无人机飞行路径代价与航迹长度、终点距离两个因素相关联。进行无人机路径规划,既要考虑航迹长度大小,也要考虑与终点距离因素。
设置第i段无人机飞行路径代价函数如下所示:
$ {\rm{min}}J = k \times \sum {{l_i}} + (1 - k) \times {l_{{\rm{goal}}}} $
(11) 式中,∑li表示从起始点到第i个位置点的所有轨迹段长度之和;lgoal表示第i个位置点到终点的距离;k为权重系数,在(0, 1)范围内。
-
结合激光传感器特性,使用DIJKSTRA算法与人工势场法混合算法进行路径规划。DIJKSTRA算法进行全局规划,从起始点出发,逐步生成下一个转移点直至终点,保证路径经过所有杆塔目标点。同时使用人工势场法进行局部优化,计算杆塔对无人机的引力作用调整规划路径。规划流程如下所示:(1)加载杆塔目标点位置与环境信息,对山区环境离散化;(2)使用DIJKSTRA算法进行全局路径规划,综合考虑代价函数最小原则以及无人机与杆塔的垂直距离w的限制;(3)使用人工势场法按杆塔先后顺序进行局部优化,计算杆塔吸引域内吸引力最小的点作为新的目标点;(4)再次使用DIJKSTRA算法进行全局路径规划,更新规划路径;(5)判断是否优化完所有杆塔目标点,若未完成则返回第(2)步;(6)输出最优规划路径。
无人机山区环境激光扫描路径规划方法研究
Research on laser scanning path planning method for UAV mountain environment
-
摘要: 为了提升山区环境里无人机进行扫描时的飞行测量有效性和飞行效率,提出了一种结合传感器测量特性的路径规划方法,以激光传感器的最大测量距离和角度、飞行效率、巡检目标位置为约束,通过求解规划方程得到飞行的轨迹,从而保证无人机山区环境巡检的安全和效率。结果表明,结合Velodyne VLP-16,RS-LiDAR-16,HDL-32E 3种激光传感器的最大测量距离d以及测量角度θ等特性进行路径规划,规划优化效果分别达到了7.58%, 11.18%, 13.33%。该研究验证了山区激光扫描路径规划方法的有效性和正确性。Abstract: Point cloud data of mountain power corridor space can be used to assist in path planning, reduce the probability of the drone flight accidents and effectively improve the efficiency and accuracy of the inspection with drones. In the process of path planning, the effectiveness of flight measurement and flight efficiency need to be taken into account. However, there is no effective way to solve this problem. In order to improve the effectiveness of flight measurement and flight efficiency, a path planning method combining sensor measurement characteristics was proposed, which was based on the maximum distance and angle of laser measurement, with flight efficiency and inspection target position as the constraints. The trajectory of flight was obtained by solving the planning equation, so as to ensure the safety and efficiency of the drone mountain environmental inspection. The experimental results show that the planning optimization effect is 7.58%, 11.18% and 13.33%, respectively, with consideration of the maximum measured distance d and measuring angle θ evaluated by the combined laser sensor of Velodyne VLP-16, RS-LiDAR-16 and HDL-32E. Therefore, the validity and correctness of the path planning method of laser scanning in mountain areas were verified.
-
Key words:
- laser technique /
- unmanned aerial vechicle /
- laser sensor features /
- path planning
-
Table 1. Laser sensor parameters
model d/m θ Velodyne VLP-16 100 ±15°, vertical RS-LiDAR-16 150 ±15°, vertical HDL-32E 100 +10°~-30°, vertical -
[1] CHEN Y, ZHOU B Zh, TAN J, et al. Research and application of airborne doppler wind lidars[J].Laser Technology, 2011, 35(6):795-799(in Chinese). [2] BLANCHARD S, KEARNEY B. Application of IRIDIUM(R) system technology to UAV based PCS services to the warfighter[C]//IEEE Military Communications Conference Proceedings MILCOM 1998.New York, USA: IEEE, 1998: 726936. [3] RAGO C, PRASANTH R, MEHRA R K, et al. Failure detection and identification and fault tolerant control using the IMM-KF with applications to the Eagle-Eye UAV[C]//Proceeding of the 37th IEEE Conference on Decision and Control. New York, USA: IEEE, 1998: 761963. [4] SMILJIAKOVIC V, GOLUBICIC Z T, MANOJLOVIC P S, et al. Application of integrated autonomous microwave position finding system and GPS for UAV navigation[C]//6th International Conference on Telecommunications in Modern Satellite, Cable and Broadcasting Service 2003.New York, USA: IEEE, 2003: 124627. [5] MITRA A K. Position-adaptive UAV radar for low-alfitude sensing applications[C]//2003 IEEE Aerospace Conference Proceedings.New York, USA: IEEE, 2003: 1235471. [6] PARK J K, AMRITA D, PARK J H. Application of agricultural subsidy inspection using UAV image[C]//Remote Sensing for Agriculture, Ecosystems, and Hydrology, Ⅹ Ⅷ.New York, USA: IEEE, 2016: 2241324. [7] ALIREZA A, PAYAM A, KANG K, et al. Neural adaptive observer-based sensor and actuator fault detection in nonlinear systems: Application in UAV[J]. ISA Transactions, 2016, 67:317-329. [8] CARNDUFF S, COOKE A. Application of aerodynamic model structure determination to UAV data[J]. The Aeronautical Journal, 2011, 115(1170): 481-492. doi: 10.1017/S0001924000006126 [9] NAIDOO Y, STOPFORTH R, BRIGHT G. Development of an UAV for search & rescue applications[C]//IEEE Africon'11.New York, USA: IEEE, 2011: 6072032. [10] WANG Y H. Agriculture robot and applications[C]//Proceedings of the 2014 International Conference on Future Information Engineering and Manufacturing Science (FIEMS 2014). Beijing: International Research Association of Information and Computer Science, 2014: 35-57(in Chinese). [11] CLAUDIA S, ANETTE E, PIERRE K. Measuring gullies by synergetic application of UAV and close range photogrammetry[J]. Catena, 2015, 132:76-82. [12] BESADA J A, BERGESIO L, CAMPAÑA I, et al. Drone mission definition and implementation for automated infrastructure inspection using airborne sensors[J]. Sensors, 2018, 18(4): 43-54. [13] LIU Ch A, YAN X H, LIU Ch Y, et al. Dynamic path planning method for mobile robot based on improved ant colony algorithm[J]. Electronic Journal, 2011, 39(5):1220-1224(in Chinese). doi: 10.1109/INM.2011.5990675 [14] SUN Sh Q, MA J. History and reality: The development process, present situation and challenges of UAV[J]. Cruise Missile, 2005 (1): 14-19(in Chinese). [15] CHUN Y, JIANG M, ZHANG H. Development status and prospect of UAV[J]. Cruise Missiles, 2005(2): 23-27(in Chinese). [16] ZHOU X M, ZHAO L B, ZHANG X L. Discussion on image processing technology and method of low altitude UAV[J]. Surveying and Mapping and Spatial Geographic Information, 2012, 35(2): 182-184(in Chinese). [17] WANG Zh H, HUANG X N, LIANG K, et al. Research on transmission line inspection system based on four-rotor UAV[J]. China Electric Power, 2012, 45(10): 59-62(in Chinese). [18] LIU Ch G, WANG X P, LIU Ch Y, et al. Three-dimensional track planning of UAV based on improved gray wolf optimization algorithm[J]. Journal of Huazhong University of Science and Technology (Natural Science Edition), 2017, 45(10): 38-42(in Chinese). [19] YAO S H. Development status and trend of UAV[J]. Electronic Fabrication, 2018(1):96-98 (in Chinese). [20] QI Sh J, JING L, WANG Y L. Summary of UAV system and its development trend[J]. Air Missile, 2018(4): 17-21 (in Chinese). [21] HU Zh H. Research on some key technologies of UAV track planning based on intelligent optimization algorithm[D]. Nanjing: Nanjing University of Aeronautics and Astronautics, 2011: 17-23(in Chin-ese). [22] BA H T. Research on UAV track planning[D]. Xi'an: Northwest University of Technology, 2006: 32-54 (in Chinese). [23] YANG G T, SU R Q, WU H, et al. Power tower inspection path planning of flying robot[J]. Journal of Huazhong University of Science and Technology (Natural Science Edition), 2015, 43(s1): 324-327(in Chinese). [24] JIA G Zh. Research on 3-D track planning of UAV based on genetic algorithm and sparse A* algorithm[D]. Nanjing: Nanjing University of Posts and Telecommunications, 2017: 55-67(in Chinese). [25] LIU Q W, LI Sh M, LI Z Y, et al. Research progress on forestry application of UAV lidar and photogrammetry[J]. Forestry Science, 2017, 53(7): 134-148(in Chinese).