摘要

中国是世界上最大的苹果生产国，苹果产量在全球总产量中占比超过50%，新疆是中国的重要苹果生产基地。2023年全疆苹果产量216万吨。病虫害防治是果园生产中的重要环节，传统施药方式往往存在农药过量施用和药液飘移的问题，农药利用率仅有20%～30%，对果园周边的土壤、水源和空气造成污染，甚至对生态系统造成破坏。因此，果园施药作业迫切需求高精度、高效率、自动化的果树冠层对靶施药机械，为此设计了一种9自由度摇臂式果园仿形作业装置，并设计了仿形对靶施药控制系统。通过控制机具仿形包络苹果树冠层，提高喷雾作业精准度、提升药液利用率、改善喷施均匀性。主要研究内容如下：

根据测量的苹果园果树参数，设计了9自由度摇臂式仿形装置，装置包含3个平移自由度与6个旋转自由度。其中仿形作业臂设计为单侧仿形结构，主要由主臂、上侧臂、下侧臂、上小臂、下小臂、水平伸缩臂、旋转机构以及驱动各作业臂的电动推杆、减速电机组成。

（2）基于拉格朗日法建立装置动力学方程，分析装置动力学特性，基于Adams建立装置动力学模型，通过分析装置在运动过程中的受力情况验证装置设计的合理性。同时建立仿形作业臂各关节坐标系，对仿形作业臂进行正运动学分析、逆运动学分析，通过Simulink仿真验证了运动学算法的准确性，为控制系统设计提供理论支撑。

（3）完成作业臂仿形控制系统设计。苹果树冠层仿形控制系统主要由单片机系统、有刷直流电动推杆、电机驱动模块、超声波测距模块、人机交互模块、姿态检测模块等组成，基于STM32F103ZET6设计单片机系统，并设计了直流有刷电机驱动模块作为从控制器接收来自单片机信号调节多路电动推杆伸缩。采用蓝牙模块进行上下位机通讯，基于Android Studio开发了上位机APP，实现了通讯参数设定、控制指令设置、机具作业信息显示（作业时长、各点位超声波检测值）、故障状态显示等功能。运用自适应卡尔曼滤波降低超声波传感器测量冠层参数误差。采用神经网络整定PID闭环控制算法调节电动推杆伸缩以保证各检测点位到果树距离维持在设定值。

（4）开展田间试验测定作业臂最大定位偏差与仿形时间，结果表明：超声波传感器各点位距离果树冠层0.6m、0.8m、1.0m处的作业臂最大定位偏差分别为32mm、41mm、48mm；平均仿形时间分别为43.46s/棵、26.78s/棵、15.52s/棵，满足设计要求。以平均雾滴密度、平均雾滴沉积量、药液附着率和施药变异系数作为量化指标，进行了苹果树冠层仿形对靶施药控制系统作业效果验证试验。结果表明：仿形对靶施药平均雾滴密度为72.4个/cm²，平均雾滴沉积量为2.03μL/cm²，药液附着率为49.8%，施药变异系数为26.4%。相同作业条件下，仿形对靶施药相比定距对靶平均雾滴密度、平均雾滴沉积量、药液附着率分别提升64.2%、40%、29.7%，施药变异系数下降42.5%，提升了药液利用率与施药均匀性。

Abstract

‌China is the world's largest apple producer, accounting for over 50% of global apple output. Xinjiang serves as a vital apple production base in China, with a total apple yield of 2.16 million metric tons in 2023. Pest and disease control is a critical aspect of orchard management. Traditional pesticide application methods often suffer from excessive pesticide use and spray drift, with a utilization rate of only 20%–30%, leading to pollution of soil, water, air, and ecosystem damage in surrounding areas. To address these challenges, there is an urgent demand for high-precision, efficient, and automated canopy-targeted pesticide application machinery. A 9-degree-of-freedom (DOF) rocker-type orchard profiling arm and a profiling-targeted spraying control system were developed. By controlling the profiling mechanism to envelop the apple tree canopy, the system improves spraying accuracy, enhances pesticide utilization, and optimizes spray uniformity. The main research components are as follows:

Based on the measured parameters of apple orchard trees, a 9-degree-of-freedom swing-arm profiling device has been designed. The device consists of 3 translational degrees of freedom and 6 rotational degrees of freedom. The profiling arm is configured as a unilateral structure, primarily comprising a main arm, upper arm, lower arm, upper forearm, lower forearm, horizontal telescopic arm, rotating mechanism, as well as electric push rods and geared motors driving each arm.‌

(2) The dynamic equations of the device are established based on the Lagrangian method to analyze its dynamic characteristics. A dynamic model of the device is developed using Adams, and the rationality of the design is verified by analyzing force conditions during motion. Coordinate systems are established for each joint of the profiling manipulator, followed by forward and inverse kinematic analyses. The accuracy of the kinematic algorithms is validated through Simulink simulations, providing theoretical support for control system design.

(3) The profiling control system for apple tree canopy operations integrates a microcontroller system (STM32F103ZET6) with brush-type DC linear actuators, motor drivers, ultrasonic ranging modules, HMI, and BeiDou positioning modules. A DC motor driver module serves as a slave controller to regulate actuator motions via microcontroller signals. Bluetooth enables upper-lower computer communication, with an Android Studio-developed APP enabling parameter configuration, control commands, operation data display (duration, ultrasonic values), and fault diagnostics. Adaptive Kalman filtering compensates for measurement errors caused by orchard terrain unevenness and tractor nonlinear motion. A neural network-tuned PID closed-loop algorithm maintains preset distances between monitoring points and trees through actuator adjustments.

(4) The experiment on the maximum positioning deviation of the operating arm and contouring time measurement showed that: The maximum positioning deviations of the ultrasonic sensor at distances of 0.6m, 0.8m, and 1.0m from the apple tree canopy were 32mm, 41mm, and 48mm, respectively; The average contouring times per tree were 43.46 s/tree, 26.78 s/tree, and 15.52 s/tree, correspondingly. Using quantitative indicators including average droplet density, average droplet deposition, liquid attachment rate, and spraying variation coefficient, the operational effectiveness of the contouring target-oriented spraying control system for apple tree canopies was verified. The results demonstrated that the contouring target-oriented spraying achieved an average droplet density of 72.4 droplets/cm², average droplet deposition of 2.03 μL/cm², liquid attachment rate of 49.8%, and spraying variation coefficient of 26.4%. Under identical operating conditions, compared with fixed-distance target-oriented spraying, the contouring target-oriented spraying exhibited significant improvements: 64.2% increase in average droplet density, 40% enhancement in average droplet deposition, and 29.7% elevation in liquid attachment rate, along with a 42.5% reduction in spraying variation coefficient. These results indicate substantial improvements in both pesticide utilization efficiency and spraying uniformity.

参考文献

[1]任清晨.新疆资源环境承载力评价与预警研究[D].中国矿业大学,2024.

[2]郑国光.中国气候[M].北京:气象出版社,2019.

[3]中国气象局国家气候中心.全国气候影响评价[M].北京:气象出版社,2021.

[4]国家统计局. 2024年中国统计年鉴[M].北京:中国统计出版社,2024.

[5]周良富,张玲,薛新宇,等.农药静电喷雾技术研究进展及应用现状分析[J].农业工程学报,2018,34(18):1-11.

[6]孟庆珍,刘莉.烟台市苹果产业发展现状、问题及对策分析[J].北方园艺,2024(21):115-122.

[7]孙政,赖忠晓,赵晓敏,等.渭北旱塬苹果病虫害全程生物防控技术应用效果评价[J].中国农业科学,2023,56(06):1102-1112.

[8]赵映,肖宏儒,梅松,等.我国果园机械化生产现状与发展策略[J].中国农业大学学报,2017,22(06):116-127.

[9]郑永军,陈炳太,吕昊暾,等.中国果园植保机械化技术与装备研究进展[J]. 农业工程学报,2020,36(20):110-124.

[10]何雄奎.中国植保机械与施药技术研究进展[J].农药学学报,2019,21(Z1):921-930.

[11]孙鑫.可旋转式果树仿形喷雾机构的设计[D].南京林业大学,2023.

[12]刘二阳.基于农户视角的绿色种植技术认知与采纳意愿研究[D].中国农业科学院,2020.

[13]中华人民共和国农业部.到2020年农药使用量零增长行动方案[EB/OL].https://www.moa.gov.cn, 2015.

[14]陈子文,胡宗锐,熊扬凡,等.树冠环绕式仿形对靶施药机设计与试验[J].农业机械学报,2023,39(03):23-32.

[15]周良富,薛新宇,周立新,等.果园变量喷雾技术研究现状与前景分析[J].农业工程学报,2017,33(23):80-92.

[16]王起阳.山地果园仿形喷杆喷雾机的设计与试验[D].华中农业大学,2024.

[17]何勇,肖舒裴,方慧,等.植保无人机施药喷嘴的发展现状及其施药决策[J].农业工程学报,2018,34(13):113-124.

[18]翟长远,赵春江,Ning Wang,等.果园风送喷雾精准控制方法研究进展[J].农业工程学报,2018,34(10):1-15.

[19]李龙龙,何雄奎,宋坚利,等.基于变量喷雾的果园自动仿形喷雾机的设计与试验[J].农业工程学报,2017,33(01):70-76.

[20]MAHMUD M S,ZAHID A,HE L,et al.Development of a LiDAR-guided section-based tree canopy density measurement system for precision spray applications[J].Computers and Electronics in Agriculture,2021,182:106053.

[21]MANANDHAR A,ZHU H,OZKAN E,et al.Techno-economic impacts of using a laser-guided variable-rate spraying system to retrofit conventional constant-rate sprayers[J].Computers and Electronics in Agriculture,2020,21(5):1156-1171.

[22]Llorens J,Gil E,Llop J,et al.Ultrasonic and LIDAR Sensors for Electronic Canopy Characterization in Vineyards:Advances to Improve Pesticide Application Methods[J].Sensors,2011,11(2):2177-2194.

[23]OSTERMAN T,HOCEVAR M,et al.Real-time positioning algorithm for variable-geometry air-assisted orchard sprayer[J].Computers and Electronics in Agriculture, 2013,98:175-182.

[24]Hocevar M,Sirok B,Jejcic V,et al. Design and testing of an automated system for targeted spraying in orchards[J].Journal of Plant Diseases&Protection,2010,117(2):1-79.

[25]Miranda-Fuentes,Rodriguez-Lizana,Cuenca, et al.Improving plant protection product applications in traditional and intensive olive orchards through the developmentof new prototype air-assisted sprayers[J].Crop Protection,2017,94:44-58.

[26]MAGHSOUDI H,MINAEI S,GHOBADIAN B,et al.Ultrasonic sensing of pistachio canopy for low-volume precision spraying[J].Computers and Electronics in Agriculture, 2015, 112: 149-160.

[27]Jordi Llorens,Emilio Gil,Jordi Llop, et al.Ultrasonic and LIDAR Sensors for Electronic Canopy Characterization in Vineyards: Advances to Improve Pesticide Application Methods[J].Crossref, 2011, 11(2): 2177-2194.

[28]SHEN Y,ZHU H P,et al.Development of a laser-guided,embedded-computer-controlled, air-assisted precision sprayer[J].Transactions of the ASABE, 2017, 60(6): 1827-1838.

[29]ABBAS I,LIU J,FAHEEM M,et al.Different real-time sensor technologies for the application of variable-rate spraying in agriculture[J].Sensors and Actuators A: Physical, 2020,316(40):112-265.

[30]张蕾,毛建卫.浙江农业和农村科学技术发展历史研究[M].杭州:浙江大学出版社,2020.

[31]张惠,姜春淼,张景,等.果园植保仿形喷雾技术研究现状与展望[J].安徽农业科学,2023,51(04):6-11.

[32]张建瓴,陈树军,可欣荣,等.仿形喷雾装置的设计及分析[J].现代制造工程,2006(01):120-122.

[33]霍鹏.果园仿树形喷杆施药装置设计与试验[D].河北农业大学,2022.

[34]张慧春, 南玉龙,郑加强,等.一种林木喷雾机仿形机臂及其仿形喷雾的方法[P]. 江苏：CN108244084A,2018-07-06.

[35]张疼.三位一体多功能喷雾机研制与试验研究[D].山东农业大学,2016.

[36]房开拓,周良富,尤丽华.果园仿形风送喷雾机构设计与试验[J].中国农机化学报,2022,43(03):68-74+83.

[37]张慧春,南玉龙,周宏平, 等.基于树冠表型特征的仿形变量喷雾机及仿形控制方法[P]. 江苏：CN111972380A,2020-11-24.

[38]李杰,赵纯清,李善军,等.基于CFD果园风送式喷雾机雾滴沉积特性研究[J].华中农业大学学报,2019,38(06):171-177.

[39]马嘉昕,丁羽,张玉湘萌，等.果园植保喷雾机研究现状与优化建议[J]. 南方农机,2023,54(12):196-198.

[40]兰玉斌,闫瑜,王宝聚,等.智能施药机器人关键技术研究现状及发展趋势[J]. 农业工程学报,2022,38(20):30-40.

[41]翟长远,赵春江,王秀,等.幼树靶标探测器设计与试验[J].农业工程学报,2012,28(02):18-22.

[42]曾星,俞龙,孙道宗,等.超声波测距在果树冠层三维重构中的应用[J].农机化研究,2012,12(18):148-151.

[43]俞龙,洪添胜,赵祚喜,等.基于超声波的果树冠层三维重构与体积测量[J].农业工程学报,2010,26(11):204-208.

[44]冯青春,赵春江,王晓楠,等.基于视觉伺服的樱桃番茄果串对靶测量方法[J].农业工程学报,2015,31(16):206-212.

[45]丁为民,赵思琪,赵三琴,等.基于机器视觉的果树树冠体积测量方法研究[J].农业机械学报,2016,47(06):1-10,20.

[46]徐昌.基于面阵激光雷达的果树靶标重构识别系统研究[D].华南农业大学,2021.

[47]张煜恒.基于激光点云数据的果树三维重建方法研究[D].南京林业大学,2023.

[48]高斌.基于超声探测的果园对靶变量施药系统设计[D].太原理工大学,2018.

[49]肖珂,郝毅,高冠东.果园自动变距精准施药系统设计与试验[J].农业机械学报,2020,53(10):137-145.

[50]袁鹏成,李秋洁,邓贤,等.基于LiDAR的对靶喷雾实时控制系统设计与试验[J]. 农业机械学报,2020,51(增刊1):273-280.

[51]中国农业大学.一种果园机器人主动视觉探测系统及探测方法:201911390237.4[P].2019-12-30.

[52]许蒙,王琼,徐会善.低效苹果园常用改良纺锤树形改造经验[J].西北园艺,2024,10(10):18-19.

[53]李昂.全产业链视角下苹果产业投入产出效率研究—以新疆南疆苹果主产区为例[D].阿拉尔:塔里木大学,2023．

[54]张建瓴,陈树军,可欣荣,等.仿形喷雾装置的设计及分析[J].现代制造工程,2006(01):120－122．

[55]尹翔宇.变喷杆喷雾机喷杆自动仿形系统的研究[D].南京:南京林业大学,2015.

[56]向军.果园智能仿形变量喷雾控制系统设计与试验[D].广州:华南农业大学,2020.

[57]张勤,刘丰溥,蒋先平,等.番茄串收机械臂运动规划方法与试验[J].农业工程学报,2021,37(09):149-156.

[58]崔永杰,马利,何智,等.基于最优空间的猕猴桃双臂并行采摘平台设计与试验[J].农业机械学报,2022,53(08):132-142.

[59]王伟志.积灰对光伏组件影响及清灰设备机械臂研究[D].兰州理工大学,2019.

[60]董玫君,周焕银,房鹏程.机械臂运动学和动力学的李群表述[J].科学技术与工程,2024,24(14):5872-5881.

[61]吴鹏,孟祥远,王子晔,等.基于神经动力学的冗余机械臂恒定转速比算法[J].北京理工大学学报,2019,39(10):1081-1085.

[62]訾斌,赵嘉浩,王威,等.大空间刚柔耦合机器人关键技术研究现状及发展趋势[J].机械工程学报,2024,60(23):21-42.

[63]肖正明,段俊杰,周川,等.考虑连杆和关节柔性的工业机器人大臂静动态性能优化[J].农业机械学报,2024,55(07):449-458.

[64]黄鹏,张金程,丁华锋,等.一种新型装载机工作装置的运动学与动力学分析[J].机械工程学报,2024,60(01):1-13.

[65]强红宾,杜亮亮,康绍鹏,等.具有解析解的1T2R重载并联稳定平台运动学及动力学建模[J].兵工学报,2024,45(11):3959-3969.

[66]宁会峰,鄢志彬,程荣展,等.光伏组件清扫机械臂动力学建模与分析[J].太阳能学报,2020,41(12):138-145.

[67]覃艳明,赵静一,仝少帅,等.八自由度机械臂位置运动学模型解析解[J].农业机械学报,2019,50(01):400-405.

[68]梁相龙,姚志凯,邓文翔,等.六自由度液压机械臂运动学标定和逆解研究[J].机械工程学报,2025,61(02):346-357.

[69]李剑,刘贤松,张英,等.通用型多功能6自由度串联机械臂运动学综合仿真平台的设计与开发[J].机械工程学报,2024,60(23):102-113.

[70]万欢,欧媛珍,管宪鲁,等.无人农机作业环境感知技术综述[J].农业工程学报,2024,40(08):1-18.

[71]苑黎明,张宝强,姜浩,等.基于XGBoost的三维温盐反演模型声速仿真应用[J].海洋科学进展,2023,41(03):520-539.

[72]杨洋,韩华宇,安东,等.基于激光雷达的果树智能修剪系统设计与试验[J]. 农业机械学报,2024,55(07):47-56.

[73]梅闯,冯贝贝,马丽亚,等.疏密改造对新疆富士苹果果园微环境和光谱特性的影响[J]. 新疆农业科学,2024,61(07):1682-1688.

[74]国家卫生健康委疾病预防控制局.中国居民营养与慢性病状况报告[M].北京: 人民卫生出版社,2020.

[75]东南大学公共卫生学院.东南大学张徐军教授：职业性农药中毒现象亟待关注[EB/OL].https://gw.seu.edu.cn/,2015.

[76]Zaman Q U,Schumann A W,Hostler H K.Quantifying sources of error in ultrasonic measuremets ofcitrus orchards[J].Transaction of the ASAE, 2007,23(04):449-453.

[77]黄培奎,张智刚,罗锡文,等.农机具姿态倾角测量系统设计与试验[J].农业工程学报,2017,33(22):9-16.

[78]高潇,王庆辉,薛智慧,等.PSO-Fuzzy-PID复合控制器在喉栓调节发动机压强控制中的应用[J].推进技术,2024,45(07):1-10.

[79]彭治文,陈晓亮,朱珈辰,等.基于PSO-BP神经网络的游泳池式反应堆堆芯功率调节系统优化研究[J].核动力工程,2024,45(04):173-180.

[80]中华人民共和国工业和信息化部.JB/T 9782-2014植保机械通用试验方法[S].北京:机械工业出版社, 2014.

[81]中华人民共和国农业部.NY/T 650-2013喷雾机（器）作业质量[S].北京:中华人民共和国农业部,2013.