基于状态预测的田间机-地传感器系统协同采集方式研究  

Cooperative Wind Field Data Acquisition Based on Unmanned Aerial Vehicle Flight Status Modelling for Agricultural Chemical Applications

作　　者：李继宇[1,2] 兰玉彬 施叶茵[3] 张亚莉 欧阳帆 陈盛德[1,2] LI Jiyu;LAN Yubin;SHI Yeyin;ZHANG Yali;OUYANG Fan;CHEN Shengde(College of Engineering, South China Agricultural University, Guangzhou 510642, China;National Center for International Collaboration Research on Precision Agricultural Aviation Pesticide Spraying Technology, Guangzhou 510642, China;Department of Biological Systems Engineering, University of Nebraska-Lincoln, Lincoln NE 68583, USA)

机构地区：[1]华南农业大学工程学院,广州510642 [2]国家精准农业航空施药技术国际联合研究中心,广州510642 [3]内布拉斯加大学林肯分校生物系统工程系,林肯ne68583

出　　处：《农业机械学报》2018年第6期246-253,277,共9页Transactions of the Chinese Society for Agricultural Machinery

基　　金：国家重点研发计划项目(2017YFD0701001);国家自然科学基金项目(31771682);广东省重大科技计划项目(2017B010116003)

摘　　要：基于旋翼无人机的低空、低速、利用旋翼风场作业等飞行特征,采用机载北斗定位系统获取精准机体实时观测值,协同地面风速传感器构成机-地传感器采集系统,尝试对具有稳定飞行轨迹的无人机进行状态预测。在充分讨论飞行状态的预测策略、可预测性、起始点确定等问题基础上,建立状态预测模型,设计状态预测算法用以自动判定传感器采集时段的起始点。依据算法展开冠层风速田间采集试验,对于无人机预测状态数据和实际观测数据做了对比分析,发现在可置信度为99%水平时,两者无差异的概率P值为0.956;同时统计出X、Y、Z向风速最大值出现时刻均值分别为3.036、2.427、3.145 s,计算出对应的标准差分别为0.79、0.87、0.98 s,说明3向风速最大值出现时刻在5 s采样范围内具有较明显的区域性,验证了采集时刻的准确性,表明机-地协同实时采集旋翼风场数据的有效性得到了显著提高。To study the wind field pattern created by unmanned aerial vehicles （UAVs） in agricultural chemical applications, triggering the wind speed sensors distributed in crop canopy along the flight path simultaneously when the UAV passes over each of them is critical in capturing the instantaneous wind field data. However, in many cases the measurements were triggered manually by human vision which reduced the timeliness and validity of the data. The data acquisition triggering method was improved and automated for wind speed sensors by predicting the exact UAV flyover timing with accurate geo-location information from an onboard Beidou positioning system and the modeling of future flight status based on past flight data given that agricultural UAVs usually operate at low speed and low altitude without overload. Since the weed speed sensors used could only record data for five seconds, a flight status prediction model was developed to determine the triggering timing for data acquisition based on the consistency and stability of the flight direction, speed, and altitude within a certain period of time. Extensive field experiments were conducted, and the model predicted and wind speed sensor measured maximum wind speed data were compared. No significant difference was found between them at a 99% confidence interval with a P-value of 0. 956. With the improved triggering timing, the averaged maximum wind speed in X, Y, Z axes occurred at 3. 036 s, 2. 427 s and 3. 145 s, respectively, of the five-second logging period with standard deviations of 0. 79 s, 0.87 s and 0.98 s, respectively. The maximum wind speed, which corresponded to the wind speed when the UAV flew over each sensor, measured by the improved data acquisition system was ensured to be captured now within the five-second optimal logging period of the wind speed sensors by the improved aerial-and-ground-sensor cooperative sensing system.

关 键 词：农业植保 北斗定位系统 无人机 传感器 机-地协同 状态预测 

分 类 号：S252.3[农业科学—农业机械化工程] V279.2[农业科学—农业工程]