基于机器视觉和仿真试验的蔬菜穴盘幼苗移栽关键技术研究

发布时间：2018-06-20 17:08

本文选题：穴盘苗 + 自动移栽　；参考：《吉林大学》2016年博士论文

【摘要】：目前,国内的蔬菜穴苗移栽方式以人工为主,作为育苗过程的重要环节,穴苗移栽设备的自动化、机械化实现,可减少劳动强度,提高作业效率,且保证移栽穴苗的一致性效果。随着穴盘育苗技术的日渐成熟和普及,国内自动穴苗移栽设备的发展滞后已成为蔬菜穴苗行业发展的“瓶颈”,目前尚无可以在生产实际中使用的自动移栽设备。试验中移栽机械应用到实际的移栽过程时常表现为穴苗钵体破裂,幼苗受损,致使移栽成功率不高。其主要原因是国内对穴盘钵体移栽机械基础理论研究不充分、不全面,导致对实用技术支撑不足,有些成果还处于试验阶段,制约了移栽机械的发展。本文采用机器视觉和仿真试验的手段对蔬菜穴盘幼苗移栽关键技术进行研究,旨在提高移栽效率,保证移栽效果,达到一致性移栽。首先,本文针对自动化移栽过程中穴盘定位不准确而常导致移栽效果不理想的问题,提出基于双目立体视觉的穴盘定位方法。分别从二维和三维角度进行算法研究,对穴盘进行平面和立体定位,完成穴盘中心点坐标、角度定位以及从立体的角度对穴盘表面姿态的模拟。从而达到准确指导末端执行器精准取苗的目的。其次,在实际移栽过程中,移栽钵体极易受损,而统一自动化移栽手段不能及时对受损钵体进行补救,造成移栽后穴苗生长周期不同,出现参差不齐,移栽效果不理想的现象。针对这一生产实际问题,提出了一种基于形态学图像检测方法的机械手移栽钵体判别技术,该方法通过对比每个样本的三个角度图像中钵体基质面积与基准面积的比值,对钵体受损程度进行判别,及时分类、补救,达到一致性移栽效果。再次,通过多因素试验,确定了取苗执行器外形结构是影响钵体基质移栽效果的主要因素,测试的因素包括:基质含水率,取苗执行器执行结构的外形特征,作物生长初期幼苗根系对基质抗剪性能提升的效果。并将基质完整率作为试验指标,通过试验给出夹持力与夹持量的对应关系,明确各取苗执行器执行结构的最佳夹持量,采用正交试验的方法得出主次影响因素和优水平。然后,通过ANSYS/LS-DYNA有限元软件,多因素影响的移栽测试试验分析结论,在测试了钵体基质的物理参数的同时,分析了不同末端执行器执行结构对钵体基质作用的应力响应。最后,通过二次开发测试系统软件,与机械手控制平台通信,完成总体试验,通过设计总体试验方案、步骤和试验各环节,以逐步、递进的方式对本论文研究的关键技术进行验证试验,证明了各关键技术的可行性和准确性。通过对本文基于机器视觉和仿真试验的蔬菜穴盘幼苗移栽关键技术研究得出以下结论:(1)从试验结果可以看出基于双目立体视觉机械手穴盘定位方法能够比较准确模拟出穴盘表面姿态,总体误差较小,可以满足植物工厂自动化移栽要求,可用于移栽过程中引导机械手精准定位。(2)通过形态学图像检测方法对移栽钵体效果进行判别,方法简单易行,并独立开发计算基质完整率识别系统软件用于计算基质完整率和判断样本合格与否。处理时间短,准确率高,能够匹配机械手移栽流水线作业的要求,可用于取钵效果判别。(3)试验结果表明:执行结构外形特征是主要影响因素,并确定U型形状为最佳;基质含水率对取苗结果有一定影响,含水率在50%左右为最佳;在幼苗长出第1对真叶时取苗,幼苗根系对基质抗剪性能提升的效果对取苗影响不大,本文为进一步研究和改善穴盘幼苗移栽设备的关键技术点提供参考。(4)通过ANSYS/LS-DYNA有限元软件仿真分析,结果表明Z型结构、U型结构在插入钵体基质过程中应力分布合理,其对基质扰动和挤压破坏小。对三种末端执行器结构夹取钵体基质进行仿真分析,Z型结构对钵体应力集中,钵体内部Von Mises应力变化大容易使钵体破裂。B型结构、U型结构对钵体基质内Von Mises应力变化平缓,对基质扰动和挤压破坏小,在夹取过程中可保证钵体基质完整性。验证了U型结构末端执行器的合理性,其对钵体基质扰动和挤压破坏小,在夹取过程中可保证钵体基质完整性。(5)综合性能测试试验结果表明,本文提出的基于机器视觉和仿真试验的蔬菜穴盘幼苗移栽关键技术研究方法准确、可行,各关键技术为穴盘苗自动化机械化移栽提供了技术支持。总之,本文从穴盘移栽过程中存在的问题入手,采用机器视觉技术,提出穴盘定位、取钵效果判别的技术方案。同时,通过试验分析了影响基质完整率的因素,并进一步对钵体应力响应做仿真分析。本文研究成果可以作为穴盘自动化机械移栽设计的理论基础和技术支持。
[Abstract]:At present, the domestic vegetable seedling transplanting method is mainly artificial. As an important part of the seedling raising process, the automation and mechanization of the plant transplant equipment can reduce the labor intensity, improve the efficiency of the operation, and ensure the consistency effect of the transplant seedlings. With the growing maturity and popularization of the acupoint seedling raising technique, the automatic transplant equipment is used in China. The lag of development has become "the bottleneck" of the development of vegetable acupoint seedling industry. There is no automatic transplanting equipment that can be used in production practice at present. In the experiment, the transplant machine is often used in the actual transplanting process, which is often manifested by the rupture of the pot body, the damage of the seedlings and the low success rate of transplanting. The main reason is the transplanting of the potted bowl body in China. The theoretical research of mechanical foundation is inadequate and incomplete, which leads to insufficient support for practical technology, and some results are still in the experimental stage, which restricts the development of transplanting machinery. This paper uses machine vision and simulation test to study the key techniques of vegetable seedling transplanting, aiming at improving transplanting efficiency and ensuring the effect of transplanting. First, in this paper, aiming at the problem that the location of the hole disc is not accurate in the process of automatic transplanting, the problem that the effect of transplanting is not ideal is not ideal, and the method of location based on binocular stereo vision is put forward. In order to accurately guide the end actuators to accurately pick up the seedlings, the transplant bowl body is easily damaged during the actual transplanting process, and the unified automatic transplanting means can not remediate the damaged bowl in time, resulting in the different growth cycle of the seedlings after transplanting, and uneven growth. The effect of transplanting is not ideal. Aiming at the actual problem of the production, a kind of machine hand transplanting bowl discrimination technique based on the morphological image detection method is proposed. By comparing the ratio of the base area of the bowl body to the base area in the three angle images of each sample, this method discriminating the damage degree of the bowl body, classifying and repairing in time. Again, through multiple factors test, the main factors affecting the transplant effect of the seedling matrix are determined by the multi factor test. The factors of the test include the moisture content of the matrix, the shape characteristics of the executive structure of the seedling, and the effect of the root to the shear performance of the substrate at the early stage of the crop growth. The matrix integrity rate is taken as the test index, and the corresponding relationship between clamping force and holding capacity is given through the test. The optimum holding capacity of the executive structure of each seeding actuator is clearly defined. The influence factors and the superior level of the main and secondary factors are obtained by the orthogonal test method. Then, through the ANSYS/LS-DYNA finite element software, the transplanting test analysis of multiple factors is analyzed. On the basis of testing the physical parameters of the bowl matrix, the stress response of the actuator structure of the different end actuators is analyzed. Finally, through the two development and testing system software, communication with the control platform of the manipulator is used to complete the overall test, and through the design of the general test scheme, the steps and the test links, progressively and progressively. To verify the feasibility and accuracy of the key technologies in this paper, the following conclusions are obtained through the research on the key techniques of the transplant of vegetable cavern seedlings based on machine vision and simulation test. (1) it can be seen from the experimental results that the position of the binocular stereovisual manipulator is located. The method can accurately simulate the posture of the surface of the hole plate, and the overall error is small. It can meet the plant plant automatic transplanting requirements and can be used to guide the precise positioning of the manipulator in the process of transplanting. (2) the effect of the transplanting bowl body is judged by the morphological image detection method, the square method is simple and easy, and the complete rate recognition of the matrix is independently developed. The system software is used to calculate the integrity of the matrix and judge whether the sample is qualified or not. The processing time is short and the accuracy is high. It can match the requirements of the manipulator transplanting assembly line, and it can be used to determine the effect of the bowl. (3) the results of the test show that the main influence factor is the shape of the execution structure, and the shape of the U is the best; the matrix moisture content is taken The results of seedling have a certain effect. The water content is about 50%. The effect of seedling root on the shear resistance of the seedlings is not significant. This paper provides a reference for further research and improvement of the key technical points of the transplant equipment of the acupoint seedlings. (4) through the ANSYS/LS-DYNA finite element software simulation. The results show that the stress distribution of the Z type structure and the U type structure is reasonable in the process of inserting the bowl body matrix, and the matrix disturbance and the extrusion damage are small. The simulation analysis on the matrix of the bowl body of the three terminal actuator structures is carried out, the stress concentration of the Z type structure is concentrated on the bowl body, and the change of the Von Mises stress in the bowl body makes the bowl body rupture.B type structure and U type knot. The stress change of Von Mises in the bowl matrix is slow, the matrix disturbance and the extrusion failure are small, and the matrix integrity of the bowl can be ensured during the clamping process. The rationality of the end actuator of the U type structure is verified. The matrix integrity of the bowl matrix is small, and the matrix integrity of the bowl can be ensured during the clamping process. (5) comprehensive performance test test The results show that the key technology of vegetable seedling transplantation based on machine vision and simulation test is accurate and feasible, and the key techniques provide technical support for automatic mechanization transplanting of acupoint seedlings. A technical scheme of bowl discriminatory effect. At the same time, the influence factors of matrix integrity rate are analyzed by experiment, and further to the pot body stress response simulation analysis. The results of this study can serve as a theoretical basis and technical support tray design automation mechanical transplanting.
【学位授予单位】：吉林大学
【学位级别】：博士
【学位授予年份】：2016
【分类号】：S223.94

【参考文献】