精准农业变量喷药系统建模与控制方法研究
文献类型：学位论文
| 作者 | 郭娜
|
| 学位类别 | 博士 |
| 答辩日期 | 2014-11-26 |
| 授予单位 | 中国科学院沈阳自动化研究所 |
| 授予地点 | 中国科学院沈阳自动化研究所 |
| 导师 | 胡静涛 |
| 关键词 | 旁路式喷药 注入式喷药 模糊控制 Smith预估控制 内模控制 动态矩阵控制 |
| 其他题名 | Modeling and Control of Variable Rate Spraying System for Precision Agriculture |
| 学位专业 | 机械电子工程 |
| 中文摘要 | 精准农业技术根据农田不同区域土壤养分、作物生长、病虫草害等存在的时间和空间差异性，按小区域的具体需求精细准确地调整各项作业管理措施，以最少的资源获得最大经济、环境、社会效益，实现农业生产的可持续发展，是现代化农业发展的必然方向。变量作业技术是精准农业的田间实现，根据农机所在的位置和田间的实际需求，精准地控制农机具实现变量播种、施肥、喷药、灌溉等过程。我国农业机械种类、水平等方面与国外发达国家存在较大的区别，如何根据我国广泛应用的农业机械类型和水平，开发设计适合我国国情的智能变量作业机具对促进我国农机装备向自动化、智能化方向发展，促进我国农业生产的可持续性发展具有重大的意义。 变量喷药技术是精准农业变量作业的热点研究内容之一。为实现化学药剂的田间按需喷洒，国内外学者提出了多种变量喷药系统的构成方式，并对各个系统的性能进行了详细的分析，提出了许多改进方案。但是，变量喷药系统的建模和控制方向的研究内容并不多见，控制算法仍以PID和模糊控制为主。本文以变量喷药技术为研究对象，对变量喷药系统的建模方法进行了研究。针对变量喷药系统的非线性大滞后的特点，进行了喷药量控制算法的研究以改善系统的动态性能。同时，考虑到拖挂喷药机的轮式农机行驶速度扰动问题，本文提出了适用于不同变速机构的农机行驶速度控制算法以提高变量喷药精度。 本文的主要研究内容包括： 第一，旁路节流式变量喷药系统建模与控制方法研究。首先基于流体网络理论建立了变量喷药系统的流体网络模型，该模型能够很好的描述稳态时管路中各个分支的流量分配情况，并将该模型用于Smith-模糊PID预估控制器的设计中以减小系统延迟的影响，加快了喷药量的响应速度，增加了系统的稳定性。另外，由于主管路流量不能及时反映喷头流量的变化，因此本文提出了基于喷头压力的多模型内模控制算法，建立了多个喷头压力的辨识模型，针对每个模型建立了内模控制器，有效地减小了喷药响应的超调。 第二，注入式变量喷药系统建模与控制方法研究。旁路节流式变量喷药系统作为水溶剂控制系统，药液注入系统通过调节电磁计量泵的控制频率控制药液注入流量，建立了药液注入系统的模型，并针对注入式喷药系统存在的大传输延迟问题，设计了基于药液注入量阶跃响应模型的动态矩阵算法用于喷药量的调节，实验结果表明该算法控制下系统不具有超调，控制精度高，响应速度快。 第三，变量喷药系统行驶速度控制方法研究。本文提出了变论域模糊PID速度控制算法用于拖挂喷药机的轮式农机行驶速度控制，以提高变量喷药精度。农业机械的动力传递系统以机械式或液压式为主，其数学模型十分复杂，变论域模糊控制不依赖对象的模型，论域随误差的减小或增大而相应的收缩和扩展，提高了模糊控制的鲁棒性和自适应能力，在拖拉机与插秧机实验平台上进行了系统功能与控制算法验证，结果表明变论域模糊PID控制算法能够用于具有不同变速器的农机的行走速度控制。 第四，变量喷药控制系统设计与实现。本文设计开发了基于CAN总线的通用变量作业控制器用于气吸式播种机、外槽轮式变量施肥机、喷杆式喷药机等多种农机具的变量作业控制以及拖挂农机具的农机行驶速度。以雷沃欧豹TA800拖拉机和东方红3W-250喷杆喷药机为实验平台，构建了以通用变量作业控制器为核心的旁路节流式和注入式变量喷药系统，实现变量喷药以及拖拉机行驶速度控制功能，并进行了系统整体实验验证，结果表明，旁路节流式和注入式变量喷药系统均能实现喷药量控制，满足变量喷药误差小于5%的要求。 |
| 索取号 | TP273/G95/2014 |
| 英文摘要 | Due to the considerably spatial and temporal variability in soil, yield, weed, insect and other factors that exist within a field, precision agriculture, as known as site specific crop management, applies the right product at the right rate in the right place across the field with the aim of equalizing the application of field inputs while maximizing production across the entire field. Variable rate technology combines a variable rate control system with application equipments to apply the field inputs at a precise time and location, such as fertilize according to the soil fertility, or spray the herbicide exactly at weed. Although precision agriculture have been introduced into China since early 1990s, the arable land was distributed to every rural household due to the household contract responsibility system in China, and those farms can’t afford the expensive variable rate applicators. Therefore, the variable rate technology suitable for small-scale land and complex terrain has great significance to promote sustainable development of China's agricultural production. Variable rate spraying technology has been the hot spot in precision agriculture. The researchers have presented a variety implementation method of variable rate spraying, analyzed the performance of each system, and proposed many improvements. However, the researches of modeling and control are rare. The control strategy is still PID and fuzzy control. Thus, the objective of this paper is to development compact variable rate spraying systems suitable Chinese situation, and design the proper control strategy for the nonlinear spraying system with large lag. At the same time, because the speed of spraying system is huge disturbance, this paper proposes a speed control algorithm to apply for the agricultural vehicle with different transmission to improve the accuracy of spraying. The main contents of this paper include: Firstly，Modeling and control of pressure based variable rate spraying system with bypass throttle. The variable rate spraying system model was established based on graph theory to describe the flow distribution in spray pipes, and the hydraulic resistances of the spraying graph were analyzed and determined by the matrix operation and some experiments. The main flow and nozzle pressure are used for feedback signal separately. The Smith predictor is combined with the fuzzy self-tuning PID controller for regulating the spray rate according to the main flow to reduce the impact of time delay and nonlinearity, accelerate the spraying response speed, and increase the stability of the system. Moreover, the main flow is unable to reflect the nozzle flow in time, therefore, the nozzle pressure based multiple models and internal model control strategy was developed to minimize the overshoots. Secondly, Modeling and control of direct inject spraying system. The carrier flow and chemical flow are adjusted independently and there is no coupling between those two systems. Due to the considerable transport delay that exist between the inject point concentration and nozzle concentration, the dynamic matrix control strategy is introduced for chemical injection flow control, there is no overshoot and the system has high control accuracy and fast response character. Thirdly, traveling speed control of the agricultural vehicles during the variable rate spraying. In order to improve the variable rate spraying accuracy, the speed of vehicle that towing the variable rate spraying system is controlled. The power train system is varied with the vehicle, and difficult to model, so the variable universe strategy is introduced in the designed of fuzzy-PD controller to improve the robustness and adaptability, and the test is accomplished on tractor that has mechanical transmission and rice transplanted that has hydraulic mechanical transmission. Finally, design of variable rate spraying system. The CAN based distributed variable rate control system is designed for multiple agricultural equipment to realize the site specific management in the field, and the speed control of the equipment. The 3W-250 sprayer is modified to pressure based and direct inject spraying system, and the whole equipment is mounted on LOVOL TA800 tractor to provide pump power and stable speed. The two systems meet the requirements of variable rate spray control. |
| 语种 | 中文 |
| 产权排序 | 1 |
| 页码 | 114页 |
| 源URL | [http://ir.sia.ac.cn/handle/173321/16754]  |
| 专题 | 沈阳自动化研究所_信息服务与智能控制技术研究室 |
| 推荐引用方式 GB/T 7714 | 郭娜. 精准农业变量喷药系统建模与控制方法研究[D]. 中国科学院沈阳自动化研究所. 中国科学院沈阳自动化研究所. 2014. |
入库方式： OAI收割
来源：沈阳自动化研究所
浏览0
下载0
收藏0
其他版本
除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。